1
00:00:19,370 --> 00:00:28,310
Welcome back, everybody. I am so. And thank you very much for this opportunity to say a few words at this event.

2
00:00:28,310 --> 00:00:31,790
And just in case I forget later,

3
00:00:31,790 --> 00:00:37,280
just can I wish Steve a very happy retirement and the very interesting retirement I

4
00:00:37,280 --> 00:00:44,090
find this this this word retirement rather difficult to comprehend in your context.

5
00:00:44,090 --> 00:00:49,160
And it actually reminds me of one of Dave's favourite stock phrases.

6
00:00:49,160 --> 00:00:59,550
When he was about to introduce one of his more extravagant ideas in the lab, he would always begin it by saying, I mean, I don't believe that, but.

7
00:00:59,550 --> 00:01:02,940
And I think I have been asked, by the way,

8
00:01:02,940 --> 00:01:10,710
before I go any further to say that for those people who are listening to this session online, you can ask questions if you want.

9
00:01:10,710 --> 00:01:12,450
We haven't had any questions online yet,

10
00:01:12,450 --> 00:01:20,790
but you are perfectly at liberty to ask questions at the end of the sessions by doing the raise hand thing on Zoom so you can do if you like.

11
00:01:20,790 --> 00:01:29,700
OK. And I guess I felt this picture really had to have a bubble and and it was somewhat

12
00:01:29,700 --> 00:01:33,330
in reference to the fact that one of the things I learnt very soon after coming to

13
00:01:33,330 --> 00:01:38,310
Glasgow was that it was best to hide your pens as soon as Dave came anywhere near your

14
00:01:38,310 --> 00:01:42,750
bench because serious safety would pick it up and walk off with it somewhere else.

15
00:01:42,750 --> 00:01:52,320
And we had this hypothesis that there was a huge sum of pens somewhere up in his office that we never quite got to the bottom of.

16
00:01:52,320 --> 00:01:56,040
OK, so I'm going to talk a little bit about the Glasgow years,

17
00:01:56,040 --> 00:02:03,450
and I should emphasise I was only there towards the second half of the Glasgow years, but there was there from 1980.

18
00:02:03,450 --> 00:02:09,630
And I do have some pictures of of the the IRA preceding my arrival.

19
00:02:09,630 --> 00:02:17,840
And here is Dave under the mistletoe in the genetics department library professing.

20
00:02:17,840 --> 00:02:22,080
And yeah, they look as if they're having a good time. So that's fine.

21
00:02:22,080 --> 00:02:29,790
And as we have heard already, there are some people around to who back here, Lorraine and others who were here at that time.

22
00:02:29,790 --> 00:02:34,470
The only thing that worries me slightly is the corpse in the right hand side of the picture,

23
00:02:34,470 --> 00:02:41,400
who seems to be a boy dressed as a girl who probably seems to have electrocuted himself on the high voltage parapet.

24
00:02:41,400 --> 00:02:46,320
But, you know, safety was like that in those days. It was, it was.

25
00:02:46,320 --> 00:02:52,120
So, yes, so. So.

26
00:02:52,120 --> 00:02:56,200
So I arrived in 1987, so this is this is a picture of Dave,

27
00:02:56,200 --> 00:03:06,070
actually and my mom and dad's back garden in 1988 and I arrived in Glasgow on Black Monday.

28
00:03:06,070 --> 00:03:09,850
In 1987, the day the stock market crashed,

29
00:03:09,850 --> 00:03:24,370
which I always attributed to the fact that they had just left Zenica and they had discovered that I had moved to this slab and and I should say.

30
00:03:24,370 --> 00:03:33,190
Thank you. I always wanted to thank Dave for the fact that he actually went to the trouble of offering me a job because I didn't apply it to Dave.

31
00:03:33,190 --> 00:03:40,960
But when I was coming before I came to Glasgow, I applied for a completely different job and it somewhat miraculously to me.

32
00:03:40,960 --> 00:03:44,530
Dave actually contacted me and said, Are you interested in a job that I have?

33
00:03:44,530 --> 00:03:47,710
So I arrived in a very peculiar manner,

34
00:03:47,710 --> 00:03:55,360
but I'm forever grateful that he didn't think it was a total crank being an industrial chemist wanting to do genetics.

35
00:03:55,360 --> 00:04:05,320
And so things for a career. And as soon as soon became apparent when I arrived at the institute's police

36
00:04:05,320 --> 00:04:11,710
department that I was coming into was a very peculiar place and a very unusual place.

37
00:04:11,710 --> 00:04:17,590
To say it was buzzing was an understatement. I would say all sorts of things going on.

38
00:04:17,590 --> 00:04:24,060
And it was a very remarkable place to work and quite different from anything I had been before.

39
00:04:24,060 --> 00:04:28,270
And this is the actual building where it all took place.

40
00:04:28,270 --> 00:04:31,840
This is the genetics department photograph taken a few years ago,

41
00:04:31,840 --> 00:04:38,890
and the arrows indicate the arrow on the left indicates where this office was on the 7th floor of the building.

42
00:04:38,890 --> 00:04:46,030
And after different into it, it was for a while it was my office two and then one floor below was what we call the sixth floor,

43
00:04:46,030 --> 00:04:51,040
which was where all the labs were. And you can see from the picture this is a very tall, narrow building.

44
00:04:51,040 --> 00:04:57,090
So the floors were really quite small, and so we were all pretty crammed.

45
00:04:57,090 --> 00:05:05,980
And I don't think this building had the same sort of hierarchy that you've gotten described at Sussex Warehouse.

46
00:05:05,980 --> 00:05:10,450
But what is obvious is that we were at the top, which must be good.

47
00:05:10,450 --> 00:05:16,280
So, so but you know, looking at this picture,

48
00:05:16,280 --> 00:05:26,780
it's kind of spooky to think of all the the lives that were lived in this building during the time I was there and all the.

49
00:05:26,780 --> 00:05:38,240
All the parties that happened and all the the drinking went on and all the the substances, although I don't know anything about that,

50
00:05:38,240 --> 00:05:51,440
but I know the intrigue and the infighting and the romance, lots of romance and all happening inside inside this little building.

51
00:05:51,440 --> 00:05:58,940
I think when Dave arrived in Glasgow, the genetics department was probably quite a sleepy place,

52
00:05:58,940 --> 00:06:03,290
but mainly focussing on classical approaches to genetics.

53
00:06:03,290 --> 00:06:10,940
And he sort of led them with probably considerable kicking and screaming into the molecular era.

54
00:06:10,940 --> 00:06:18,350
And his approach to science and his enthusiasm for science seemed to rapidly attract more and more people to the genetics department.

55
00:06:18,350 --> 00:06:25,670
And so it started to fill up the building start to fill up. So even when I arrived, the building was quite congested with people.

56
00:06:25,670 --> 00:06:30,350
And I think, you know, within a few years after my arrival, it was we were packed like sardines,

57
00:06:30,350 --> 00:06:35,390
and I think that helped all the interactions that were on inside the building.

58
00:06:35,390 --> 00:06:44,620
Everybody knew everybody else was doing helped by this magical paternoster lift that went up and down the centre of the building, cutting you.

59
00:06:44,620 --> 00:06:52,340
And so the boxes that you travelled from one floor to another then were open, and once they had opened the front door, you could step on and off.

60
00:06:52,340 --> 00:07:00,050
And so when you travelled from the ground floor to the top, you went through all the worlds of all different people living in the building.

61
00:07:00,050 --> 00:07:05,570
And they held had their conversations and what they were ranting about at the time and stuff.

62
00:07:05,570 --> 00:07:10,340
So it's all very interesting and. Here's the building.

63
00:07:10,340 --> 00:07:15,650
That was a photograph of latency. And so the building is being demolished, the sixth floor has gone,

64
00:07:15,650 --> 00:07:22,520
Dave's office is gone and they're taking it down from the top because they were scared in case it was going to fall over.

65
00:07:22,520 --> 00:07:27,210
If they've tried to demolish it by traditional methods. So in a few weeks, it will be gone.

66
00:07:27,210 --> 00:07:32,150
It was all wrapped up in this polythene stuff like that man, that Frenchman,

67
00:07:32,150 --> 00:07:38,570
that dust sculptures and has been is going to be demolished on the inside.

68
00:07:38,570 --> 00:07:45,650
So the sixth floor, I think I'm afraid, is rubble now. So that's that's a shame.

69
00:07:45,650 --> 00:07:50,000
So here here is one of Dave's legendary barbecues.

70
00:07:50,000 --> 00:07:56,690
I picked this photograph just because that maximise the number of people from his grip that got it.

71
00:07:56,690 --> 00:08:01,130
And I apologise for my poor photography skills. But on the left hand side, there is Stephen Bell,

72
00:08:01,130 --> 00:08:07,680
who's going to talk in a few moments and some other people who are at the meeting such time as that in there.

73
00:08:07,680 --> 00:08:11,540
And these are other people in the group at the time.

74
00:08:11,540 --> 00:08:13,420
I said I should have said, by the way,

75
00:08:13,420 --> 00:08:19,310
I'm sure that all these people who unfortunately cannot be here would wish you the best wishes as well on this occasion.

76
00:08:19,310 --> 00:08:27,020
It was a shame, of course, because of COVID that, but it was difficult to get this event organised.

77
00:08:27,020 --> 00:08:38,420
And and yeah, so so David, you've had already was an inspirational teacher and mentor to students.

78
00:08:38,420 --> 00:08:47,270
And I think the first thing to say about that was that he was so because students and people

79
00:08:47,270 --> 00:08:52,850
recognise that they're talking to a world class scientist and that inspires them by itself.

80
00:08:52,850 --> 00:08:58,760
And so I think just to have someone like that to to to talk to was a great thing.

81
00:08:58,760 --> 00:09:08,210
But another another thing that I think makes an outstanding teacher is, is your idiosyncrasies and the little things that make you an individual.

82
00:09:08,210 --> 00:09:13,250
And one of the characteristics that they've had was was that you really love using models.

83
00:09:13,250 --> 00:09:15,830
And so you would always have something in his pocket, you know,

84
00:09:15,830 --> 00:09:25,730
fill out some sort of piece of chirping like this and illustrate some aspects of the mechanism, like super calling or recombination in this case.

85
00:09:25,730 --> 00:09:37,590
And this inspired Gary Larson. So to make this captain for his 1990 power site calendar, and you can tell, I mean,

86
00:09:37,590 --> 00:09:49,520
you might think this is a fake, but it's a sign that the tops was clearly genuine. And so, so yeah, that's so characteristic post.

87
00:09:49,520 --> 00:10:00,140
And another another thing that was very capitalistic of Dave's teaching methods was his love of offering rewards for people.

88
00:10:00,140 --> 00:10:09,290
So. So if that these rewards ranged from the low value miles pass through pints of beer bottles

89
00:10:09,290 --> 00:10:14,540
of old whisky up to the the almost mythical creator of champagne that you could get,

90
00:10:14,540 --> 00:10:17,810
depending on the scale of the challenge that he was putting to you.

91
00:10:17,810 --> 00:10:25,910
And so in the next year's farm site calendar, Gary Larson came up with this character as well.

92
00:10:25,910 --> 00:10:40,130
And so I found these in my shoe box of some of the photographs and things of the time.

93
00:10:40,130 --> 00:10:45,170
The the true adulteress of these pictures shall remain completely nameless.

94
00:10:45,170 --> 00:10:48,980
I'll bet it is. I mean, I see he or she is listing.

95
00:10:48,980 --> 00:11:00,590
Perhaps so. So these two pictures actually both reflect the fact that a lot of research in this lab at the time was looking at the exit system.

96
00:11:00,590 --> 00:11:05,000
And of course, all of these options were a big deal in the mechanism of Xia.

97
00:11:05,000 --> 00:11:09,500
So the science that was going on in the time that I was there was really divided

98
00:11:09,500 --> 00:11:13,850
into three sections that were the transposition people working on and supplements,

99
00:11:13,850 --> 00:11:23,570
et cetera. There was the main the biggest part of the the research group were working on the car and I was privileged to be in the

100
00:11:23,570 --> 00:11:32,750
lab when Sun Column's tried to convince me that like a colony on one of his petri dishes was an interesting mutant,

101
00:11:32,750 --> 00:11:37,070
but turned out to be the first mutant CRC.

102
00:11:37,070 --> 00:11:40,670
And so that was all the exciting things happening at the time.

103
00:11:40,670 --> 00:11:47,030
And then there were the result of these people, and I was recruited to join the result of these people.

104
00:11:47,030 --> 00:11:54,050
So there were three of us at the time. There was embezzlers Martin, who was at the meeting and and me.

105
00:11:54,050 --> 00:11:58,760
And what I started that this was the state of our knowledge.

106
00:11:58,760 --> 00:12:06,860
But we've already had by the way of this result, this protein, because Gordon Duncan mentioned that he was the first person to see the protein.

107
00:12:06,860 --> 00:12:16,260
But by 1987. It said it was looking like this, it was a Soviet recombination, and this picture was drawn by Martin.

108
00:12:16,260 --> 00:12:22,530
And it shows what we thought the the the pathway for recombination was.

109
00:12:22,530 --> 00:12:28,170
And so we thought that the two, the two transposons sites which are recognised by there's always protein,

110
00:12:28,170 --> 00:12:35,430
they're brought together into this thing, which is inside the square box, which we called the synapse where the DNA is all intertwined.

111
00:12:35,430 --> 00:12:45,210
And then in the final step, the strands are broken at the triangles and exchanged to make wood products, which are through DNA circles.

112
00:12:45,210 --> 00:12:52,680
And that process was thought to happen by something that was equivalent to rotating the DNA end through 360 degrees.

113
00:12:52,680 --> 00:12:57,710
And so this this very basic picture was where we were at in 1987.

114
00:12:57,710 --> 00:13:05,760
But I think I've spent the next 30 years of my career basically making this more elaborate and making it more three dimensional.

115
00:13:05,760 --> 00:13:16,530
But essentially, this picture was right away back then. And so here's what we think where we can now have the the subunit rotation mechanism.

116
00:13:16,530 --> 00:13:20,630
We think the DNA is rotating like this and we have structures to back that up.

117
00:13:20,630 --> 00:13:29,880
And in very recently, we've been doing single molecule experiments using fluorescent little for labels where we can see the

118
00:13:29,880 --> 00:13:36,000
strand exchange going from low fat and on recombinant Typekit recombinant over a period of seconds.

119
00:13:36,000 --> 00:13:39,460
So we can see these things happening in real time.

120
00:13:39,460 --> 00:13:49,030
And then also, we have the this is now a three dimensional model of the set up structure, which is hopefully going to get published very soon.

121
00:13:49,030 --> 00:13:53,960
And finally, after all the models which were proposed for this snap structure, we finally think we've got one.

122
00:13:53,960 --> 00:14:00,340
That's right. There are many models that seemed wrong even when they were published in high profile journals.

123
00:14:00,340 --> 00:14:09,490
But we think this one million is what it really looks like. And remarkably, the authors of this paper, which we're hearing at the moment,

124
00:14:09,490 --> 00:14:16,540
I think fight five of the authors were people who were at Glasgow in your time Martin,

125
00:14:16,540 --> 00:14:21,430
Sally, Mary Burke, myself, and also Alistair MacDonald, who was an undergraduate.

126
00:14:21,430 --> 00:14:27,490
I think at the time are all going to be authors on this paper, so it still has a connexion back to the days when Dave was there.

127
00:14:27,490 --> 00:14:34,510
I just wanted to get a picture of Mary on this slide. Just to mention, Mary was the technician in Dave's lab throughout his time in Glasgow.

128
00:14:34,510 --> 00:14:39,160
And then I inherited Mary after Dave left until her retirement.

129
00:14:39,160 --> 00:14:45,640
She was a fantastic technician. She made a huge contribution to all our work, as well as I'm sure many of you in the audience know,

130
00:14:45,640 --> 00:14:53,140
contributing to make sure that the lab was kept and the people were kept in order and everything, everything went smoothly.

131
00:14:53,140 --> 00:15:00,610
So thank you very much, Mary as well. So I think I've probably missed out of things, but I think I'd better stop there.

132
00:15:00,610 --> 00:15:05,320
So thank you very much and we'll move on to our message.

133
00:15:05,320 --> 00:15:13,410
I think from Colin Stirling. Is that correct? Hi, Dave, and hello, everyone.

134
00:15:13,410 --> 00:15:23,110
I'm really sorry that I'm not able to get out of Australia to come there to join you and help celebrate this your marvellous scientific career.

135
00:15:23,110 --> 00:15:29,460
I've seen the programme and it looks spectacular and it would have been great to spend a couple of days immersed in science,

136
00:15:29,460 --> 00:15:34,590
something that I, as the vice chancellor, I don't get to do very often, especially in a pandemic.

137
00:15:34,590 --> 00:15:38,940
And it'd be great to just have all the excitement of cutting edge science again.

138
00:15:38,940 --> 00:15:49,380
So I wish you all the best for that. It's a long time since I was in the lab because I was a young Ph.D. student way back in the mid 80s,

139
00:15:49,380 --> 00:15:57,510
and I remember the time you turned 40, and I remember thinking that 40 was pretty old back then.

140
00:15:57,510 --> 00:16:02,660
And then I realised that nearly 40 years has passed. So no, of course not at all.

141
00:16:02,660 --> 00:16:12,060
A different perspective of of time. And I realise now that you were just a young whippersnapper when you were head of department back then.

142
00:16:12,060 --> 00:16:22,860
But if I remember one major thing, actually, it was about your nearly boyish enthusiasm for science,

143
00:16:22,860 --> 00:16:27,570
your excitement about the wonder of all things scientific.

144
00:16:27,570 --> 00:16:33,690
It was infectious. It was exhilarating to be around and it was inspiring.

145
00:16:33,690 --> 00:16:43,410
I think I also took away from working in the lab the importance of doing science of of not just grinding out papers with incremental results,

146
00:16:43,410 --> 00:16:47,280
but seeking the joy of real discovery.

147
00:16:47,280 --> 00:16:56,050
And I see from your incredible list of publications about something that you've done time after time, after time.

148
00:16:56,050 --> 00:17:01,800
I also remember you telling me back in Glasgow that you thought I was getting increasingly difficult

149
00:17:01,800 --> 00:17:08,070
to get funding to do fundamental research and equally because some people had all been discovered.

150
00:17:08,070 --> 00:17:18,930
And I guess that looking at what you've done since then, you've proved yourself and then wrong and had this hugely successful career.

151
00:17:18,930 --> 00:17:27,310
So I wanted to thank you. Thank you for being such a great supervisor and thank you for the opportunities that I had in your lab.

152
00:17:27,310 --> 00:17:35,130
And thank you for enabling allowing me to cut my scientific teeth on my scientific and academic career.

153
00:17:35,130 --> 00:17:40,620
Thanks also for being such a great mentor throughout. I wish you all the best.

154
00:17:40,620 --> 00:17:49,590
I wish you and Lydia all the best in the future. I wish you much health and happiness as you enjoy whatever it is you go on to do now.

155
00:17:49,590 --> 00:17:59,510
I'm sure you'll have fun. So congratulations again on a marvellous career that brings Dave such.

156
00:17:59,510 --> 00:18:09,560
So our next speaker is Stephen Bell, who is not against Stephen, started a year after I arrived at Glasgow and did his Ph.D. there,

157
00:18:09,560 --> 00:18:17,390
and now he's at the University of Indiana University in Bloomington, and he's going to talk about chromosome architecture.

158
00:18:17,390 --> 00:18:21,680
I think there's a pun somewhere in there. Thanks, Marshall.

159
00:18:21,680 --> 00:18:28,820
Thank you, Dave. And I start with a photograph of this was.

160
00:18:28,820 --> 00:18:37,640
Can believe it, 30 years ago, we went climbing in the White Mountains in New Hampshire en route to FASEB recombination meeting.

161
00:18:37,640 --> 00:18:40,760
I'm sure I learnt a lot about recombination at the meeting,

162
00:18:40,760 --> 00:18:48,350
but I certainly learnt that jet lag and copious quantities of American beer really don't mix terribly well.

163
00:18:48,350 --> 00:18:54,710
You can tell us quite a few years ago now because Dave hadn't started dying is here that interesting blonde colour that it is now?

164
00:18:54,710 --> 00:19:00,560
And as you saw from the photo that Marshall showed, my hair actually existed back then as well.

165
00:19:00,560 --> 00:19:08,570
So I'd like to go back a couple more years than this, though, and just really echo the sentiments that we heard from from various speakers that Dave's

166
00:19:08,570 --> 00:19:13,460
teaching and his influence because I entered university fully intending to become a chemist.

167
00:19:13,460 --> 00:19:16,970
And I did molecular biology and sounded vaguely interesting.

168
00:19:16,970 --> 00:19:23,710
And it was a revolutionary moment when I went to see these electrons by Dave on bacterial gene regulation.

169
00:19:23,710 --> 00:19:30,980
Very it taught us the scenario and then showed us the data and described the experiments and got us to interpret

170
00:19:30,980 --> 00:19:37,070
the data and put them together and essentially trying to find her the lack of protein work from first principles,

171
00:19:37,070 --> 00:19:44,450
from experimental data. And that, for me, was the first time I felt that thrill of taking simple experimental results and

172
00:19:44,450 --> 00:19:49,910
putting them together and synthesising it into this entity that became the like open.

173
00:19:49,910 --> 00:19:56,630
It was utterly thrilling at work in me, something that wasn't there before, and that was it was a molecular biologist thereafter.

174
00:19:56,630 --> 00:20:02,700
So they've changed, of course, my life. Thank you. Thank you.

175
00:20:02,700 --> 00:20:09,840
So I continued to work in prokaryotic systems as I set up my career to compete with Dave,

176
00:20:09,840 --> 00:20:14,390
obviously, and say they've got a safe niche and a path less well trodden.

177
00:20:14,390 --> 00:20:15,720
That was the study archaea.

178
00:20:15,720 --> 00:20:26,820
These are organisms of the genus so colobus they grow in hydrothermal hot springs places like Yellowstone National Park 8°C, too.

179
00:20:26,820 --> 00:20:32,880
And the reason we study them and study them is because, well, principal justification is because of their theology.

180
00:20:32,880 --> 00:20:41,100
They exist in the information processing machineries. They use the same machineries that our cells use to affect gene transcription DNA replication.

181
00:20:41,100 --> 00:20:47,370
But in a simplified ancestral form and then a heat stable for making biochemistry almost fun.

182
00:20:47,370 --> 00:20:53,730
And this just lists their struggle against proteins that carry out replication in archaea and eukaryotes

183
00:20:53,730 --> 00:20:59,340
so that the machinery works in the context of what seems like a rather eukaryotic cell cycle as well.

184
00:20:59,340 --> 00:21:03,420
We have got phases of G1 phase, extensive G2 phase.

185
00:21:03,420 --> 00:21:10,530
We have multiple origins of replication per chromosome. We showed this years ago to all three fire once per sale for cell cycle to fire synchronously.

186
00:21:10,530 --> 00:21:15,720
One a little bit later after the completion of replication is just a chromatin are held together

187
00:21:15,720 --> 00:21:20,670
for the majority of this extended two period of the cell cycle prior to cell division.

188
00:21:20,670 --> 00:21:25,890
And we call it is not mitosis, but we call it phase.

189
00:21:25,890 --> 00:21:32,040
And cell division is one of these interesting things, because while a lot of our work was based on our soldiers conservation,

190
00:21:32,040 --> 00:21:36,180
there's also some interesting things missing in Sulphur Lewis and his kin.

191
00:21:36,180 --> 00:21:40,500
And one of those interesting things on lupus heart. That's FDIC.

192
00:21:40,500 --> 00:21:48,400
It's FDIC is nearly universal across prokaryotes, archaea and bacteria effecting cell division with the cranor katavich.

193
00:21:48,400 --> 00:21:53,110
So for lupus belongs like FDIC, they divide. How do they do it?

194
00:21:53,110 --> 00:21:57,340
And so just following molecular detective, whatever we found, purely curiosity.

195
00:21:57,340 --> 00:22:04,830
What is your faith in cell division in these organisms? And we discovered our kale counterparts of the eukaryotic S-Corp machinery.

196
00:22:04,830 --> 00:22:06,150
And this is the very same machinery.

197
00:22:06,150 --> 00:22:16,020
It carries a membrane, a decision in our cells that the final stages of cytokinesis present in a minimal ancestral form in the archaeal cells.

198
00:22:16,020 --> 00:22:27,600
Another conundrum presented itself in the SME proteins, and we've heard of it Red 50, the universal, absolutely ubiquitous DNA repair protein,

199
00:22:27,600 --> 00:22:32,910
but one thing that was obviously lacking in the FluMist genome and other craner care was

200
00:22:32,910 --> 00:22:40,180
condensing conventions universal otherwise and yet manifestly lacking in this health globally.

201
00:22:40,180 --> 00:22:44,730
So we reasoned that perhaps as an equal over most beef, as a lineage specific acquisition,

202
00:22:44,730 --> 00:22:52,260
perhaps there was some sort of lineage specific compensation for the loss of condensing in the ancestral form.

203
00:22:52,260 --> 00:23:03,330
And so we identified some novel SNC superfamily proteins and one that we identified in the social values we're going to call coalesced.

204
00:23:03,330 --> 00:23:07,740
Coalescence, a rather small some sea proteins, only sixty eight killer dolphins.

205
00:23:07,740 --> 00:23:16,740
It has, instead of the classic hinge dimerisation domain that condensed and has a has cysteine residues of putative zinc hook,

206
00:23:16,740 --> 00:23:23,220
a feature that we just had about four marines present in the rad 50 proteins and we've now identified some partner proteins.

207
00:23:23,220 --> 00:23:31,880
I wouldn't talk about them today. We've gone on to solve the crystal structure of most of the protein, we have EPAs heads,

208
00:23:31,880 --> 00:23:41,960
just classic SMS heads call Carlos coming out forms of Homo dimer via dimerisation interface at the other end of the coil called phylogenetically,

209
00:23:41,960 --> 00:23:46,460
it does look like it's clustering closest to We are killed at 50 14,

210
00:23:46,460 --> 00:23:54,840
so we seem to have an odd lineage specific adaptation of around 50 like Putin to do whatever this thing does.

211
00:23:54,840 --> 00:23:59,200
This is the dimerisation interface zoomed in on, and indeed it is a zinc.

212
00:23:59,200 --> 00:24:07,980
So this is clearly a very rad 50 like protein, a single zinc coordinated by four systems to try and work out what the protein was doing.

213
00:24:07,980 --> 00:24:14,460
It's essential we couldn't all, but we just kind of chip seek experiments to see what where was it interacting with in the chromosome?

214
00:24:14,460 --> 00:24:23,670
And so a very interesting non random distribution. We also noted that whatever cool lesson was high, transcription was really low.

215
00:24:23,670 --> 00:24:33,460
So it's an anti correlation between pool and occupancy and transcription adjudge by RNA polymerase chips are an easy sell.

216
00:24:33,460 --> 00:24:40,660
So we wondered, has it got something to do with chromosome confirmation, but of course, nobody is working on chromosome confirmation.

217
00:24:40,660 --> 00:24:44,260
So we have to establish what that confirmation itself might be.

218
00:24:44,260 --> 00:24:50,650
So we adapted chromosome confirmation capture techniques and performed them on self-focus.

219
00:24:50,650 --> 00:24:55,090
And these are just the all contact match where we plot interaction of every Lucas across

220
00:24:55,090 --> 00:24:59,110
the chromosome with every other locus across the chromosome in the second dimension,

221
00:24:59,110 --> 00:25:05,380
giving you these plots. And so we can process them to generate these Pearson correlation coefficient heat maps.

222
00:25:05,380 --> 00:25:15,700
I'm Scottish. It's tartan. I was thrilled. I was even more thrilled because it was very distinct from the patterns we see in most bacteria,

223
00:25:15,700 --> 00:25:20,380
where we have this beautiful cancer zipping of the arms of the chromosome and rather more

224
00:25:20,380 --> 00:25:27,130
reminiscent of the compartmentalised architecture that we see in method's own chromosome.

225
00:25:27,130 --> 00:25:32,970
We have large domains along the primary axis, and we have domains interacting in higher order space.

226
00:25:32,970 --> 00:25:37,440
We can analyse this using principal component analysis to determine who interacts with

227
00:25:37,440 --> 00:25:42,810
whom and define the mean boundaries and so we can classify in a compartment and a B

228
00:25:42,810 --> 00:25:50,370
compartment and where we came back to Colson was at this point when we realised that the

229
00:25:50,370 --> 00:25:55,950
B compartment exquisitely lines up with these regions of enhanced coalition occupancy.

230
00:25:55,950 --> 00:26:02,850
Even within the B compartment, the coalition isn't universally distributed if there's still irregularities within there.

231
00:26:02,850 --> 00:26:09,690
So we'll delve into that deeper in a minute or two. Again, we could see, as we've shown already,

232
00:26:09,690 --> 00:26:14,910
that the just qualitatively one can see the bee compartment is transcribed at a lower

233
00:26:14,910 --> 00:26:21,750
level because lie gives rise to lower level transcripts than does the compartment.

234
00:26:21,750 --> 00:26:30,540
And is there a causality relationship here? And so to summarise a large amount of data, we were able to show that if we raised cholesterol levels,

235
00:26:30,540 --> 00:26:34,950
we dropped inscription and reinforced at the same time the B compartment.

236
00:26:34,950 --> 00:26:39,030
If we locally or globally altered transcription, we perturbed cool s,

237
00:26:39,030 --> 00:26:49,300
and they're mutually exclusive interaction as an antagonism between these two processes on the chromosome with Lucas specific and globally.

238
00:26:49,300 --> 00:26:56,650
So at this point, we could see that we have this utterly unanticipated compartmentalised architecture of the chromosome and E compartment,

239
00:26:56,650 --> 00:26:58,300
a B compartment, a compartment,

240
00:26:58,300 --> 00:27:04,570
elevator transcription, the essential genes and the replication origins are housed within the compartment in the B compartment,

241
00:27:04,570 --> 00:27:11,080
while lower transcription coalescing. Is there non-essential genes and lots of transposons sniffing a little smelling a little

242
00:27:11,080 --> 00:27:16,480
bit like you chromatin heterochromatin if you really want to stretch an analogy.

243
00:27:16,480 --> 00:27:19,690
More recently, we've gone to investigate a higher resolution,

244
00:27:19,690 --> 00:27:25,490
so we modified the set and further use different enzymes, and we now have to kill a resolution contact map.

245
00:27:25,490 --> 00:27:31,140
So really gene and opera on level contact maps of every Lucas across the chromosome.

246
00:27:31,140 --> 00:27:35,400
We still see compartmentalisation, of course, but we see a wealth of additional features.

247
00:27:35,400 --> 00:27:37,260
And I'll just take you through those.

248
00:27:37,260 --> 00:27:45,520
So in the next slide, I'm going to take this diagonal here to strategic forty five degrees there and present it horizontal.

249
00:27:45,520 --> 00:27:51,730
And it's a very northern Indiana landscape, we see lots of mountains appearing.

250
00:27:51,730 --> 00:27:54,430
So we see domains arising along this primary diagonal,

251
00:27:54,430 --> 00:28:00,100
and we can quantify these domains using a directional preference float, basically slider window along the chromosome.

252
00:28:00,100 --> 00:28:03,410
There's the look is prepared to inject the left or to the right.

253
00:28:03,410 --> 00:28:10,510
And so that's quantified, then the lefferts interaction rate words interaction to define these domains and the boundaries between them.

254
00:28:10,510 --> 00:28:11,890
So we've got lots of small domains.

255
00:28:11,890 --> 00:28:21,290
We'll call them chromosome interaction domains and importantly, disband the entire chromosome, but in the compartment and B compartment.

256
00:28:21,290 --> 00:28:26,510
We called them chromosome interaction twins, or SIDS, because similar features have been reported initially by microbes,

257
00:28:26,510 --> 00:28:32,330
live in studies of call-backs and are now known to be a universal feature of bacterial chromosomes.

258
00:28:32,330 --> 00:28:37,790
We see these kids all along the call about a chromosome and makes love noted that boundaries between

259
00:28:37,790 --> 00:28:44,480
SIDS often correlated and demonstrate were causally linked to highly transcribed transcription units.

260
00:28:44,480 --> 00:28:46,940
We wondered if the same was true in cell for lupus.

261
00:28:46,940 --> 00:28:55,280
So we simply quantified transcripts of levels either with at said boundaries or internally in the compartment or the B compartment.

262
00:28:55,280 --> 00:29:02,050
And in both cases, you can see there's an enrichment for highly transcribed genes at boundary proximal loci.

263
00:29:02,050 --> 00:29:11,800
Note, however, that the B compartment boundary loci actually opened at a lower level than the internal transcripts in the compartment,

264
00:29:11,800 --> 00:29:17,170
so it's relatively high transcription, not absolute transcript levels as defining the boundary.

265
00:29:17,170 --> 00:29:24,970
It also suggests that perhaps it's in the B compartment and may be an additional component is imparting identity to the SIDS.

266
00:29:24,970 --> 00:29:29,560
And of course, I've told you already, the obvious candidate is coolest.

267
00:29:29,560 --> 00:29:33,970
And indeed, we see this non-random distribution of coalescing within the B compartments.

268
00:29:33,970 --> 00:29:39,250
And when we know aligner CID boundary elements onto the coolest occupancy profiles,

269
00:29:39,250 --> 00:29:46,070
we see the boundaries between SIDS are actually found areas of low correlation occupancy.

270
00:29:46,070 --> 00:29:53,620
OK, so the idea that we would have said this is the bundle region by this cool s and protein boundaries and strengthened

271
00:29:53,620 --> 00:30:03,550
by the presence of these transcripts in another little unit of folding or compaction again effected by coalition.

272
00:30:03,550 --> 00:30:05,650
Experiments are nice as well as observation,

273
00:30:05,650 --> 00:30:14,110
so we introduced the transgene into the chromosome and put in an R avenues inducible gene into an eight compartment,

274
00:30:14,110 --> 00:30:21,610
Lucas and we can switch on transcription in the presence of other avenues, have a basal level with sucrose.

275
00:30:21,610 --> 00:30:26,440
The chromosome conformation capture, and we can see that the look is where we introduced the transgene,

276
00:30:26,440 --> 00:30:29,740
even at the basal level, we see a weak said bone being generated,

277
00:30:29,740 --> 00:30:38,280
which is significantly strengthened in the presence of a ravenous inscription, is able to create a boundary element within the compartment.

278
00:30:38,280 --> 00:30:43,830
A collateral benefit of this experiment was in addition to switching on this revenues inducible transgene,

279
00:30:43,830 --> 00:30:47,310
we also switched on the endogenous arrived and it was the regulation,

280
00:30:47,310 --> 00:30:55,110
including a subtle offer on here this Distronic transcription unit, which fortuitously resides within the bee compartment.

281
00:30:55,110 --> 00:31:02,550
And then the B compartment to upon induction with transcription, we see generation of a novel set boundary again showing that transcription is

282
00:31:02,550 --> 00:31:10,550
important for defining boundary elements in both the foreign compartments.

283
00:31:10,550 --> 00:31:20,240
An additional feature that we saw on the high resolution maps was the presence of point to point loops on the chromosome synapses are at.

284
00:31:20,240 --> 00:31:25,490
Conjunction of two distinct loci, and it's about sixty five loops per genome,

285
00:31:25,490 --> 00:31:32,210
and they range massively in size loss of short range loops, but also some extremely long range loops of up to half the chromosome.

286
00:31:32,210 --> 00:31:41,390
And this is just the aggregate plots of these loop features that we can detect. So that entire half chromosome can be brought can be looped out.

287
00:31:41,390 --> 00:31:45,080
I mean, we look to see what genes were present at the anchors of these loops.

288
00:31:45,080 --> 00:31:48,860
We were really excited to know that the long range loops,

289
00:31:48,860 --> 00:31:57,140
almost every single one of them involved regions involved in ribosomal biogenesis, ribosomal RNA genes, ribosome or protein genes.

290
00:31:57,140 --> 00:32:02,900
Even some ribosomal RNA processing genes are involved in these loop anchor elements.

291
00:32:02,900 --> 00:32:09,290
So this our operating hypothesis at the moment is that we're seeing the generation of a sort of hub like organisation,

292
00:32:09,290 --> 00:32:13,460
a factory for building ribosomes and these simple prokaryotic cells,

293
00:32:13,460 --> 00:32:19,640
allowing the coordination of synthesis of ribosome, proteins, ribosome Milanese and indeed,

294
00:32:19,640 --> 00:32:25,760
the of these loop features we showed was dependent both on transcription and in cell growth phase.

295
00:32:25,760 --> 00:32:28,910
We starve the cells for ribosome will demand this law.

296
00:32:28,910 --> 00:32:36,570
These loops disappear, so it really does look like it's a coordinating organising hub and for testing that further.

297
00:32:36,570 --> 00:32:43,050
So it included this part of the talk as I showed you a compartment and being compartment superimposed upon that,

298
00:32:43,050 --> 00:32:47,100
we have these features in the compartment.

299
00:32:47,100 --> 00:32:49,500
These are bounded by highly transcribed genes.

300
00:32:49,500 --> 00:32:55,890
And so one can imagine that the genes with a couple of transcription translation might be serving as barriers to topological diffusion,

301
00:32:55,890 --> 00:33:00,690
allowing the definition of these domains between them in the B compartments.

302
00:33:00,690 --> 00:33:03,240
Well, luckily, my transcription seems to be important,

303
00:33:03,240 --> 00:33:12,110
but internal binding of core lesson seems to be important in helping to reinforce and established this architecture.

304
00:33:12,110 --> 00:33:19,910
And finally, we have the loops where we see Lucy principally within the compartment being held together even at very,

305
00:33:19,910 --> 00:33:24,240
very long ranges in the linear order of the chromosomes.

306
00:33:24,240 --> 00:33:29,160
And this is quite pleasing to see, because our previous work had shown that by modulating core lesson levels,

307
00:33:29,160 --> 00:33:34,790
we could modulate the degree of compassion and the identity of this B compartment.

308
00:33:34,790 --> 00:33:41,870
And we weren't sure if the compartment was actually being actively structured or was it simply being determined by exclusion from the B compartment,

309
00:33:41,870 --> 00:33:42,320
but no,

310
00:33:42,320 --> 00:33:53,200
the presence of these loop features provides essentially cross braces that could exist and help to impart integrity to the overall compartment.

311
00:33:53,200 --> 00:33:59,530
So that's kind of a quick run through the natural history of these chromosomes.

312
00:33:59,530 --> 00:34:02,830
And, you know, the future of birdwatching, if you like on the chromosome,

313
00:34:02,830 --> 00:34:07,660
but we're interested also in trying to determine what impact this generation of these features

314
00:34:07,660 --> 00:34:13,960
might have on the physiology of the organism and its ability to adapt to different circumstances.

315
00:34:13,960 --> 00:34:19,210
And so the final few minutes, I'll just tell you a short story and I apologise.

316
00:34:19,210 --> 00:34:25,870
Already, it's an incomplete story, but we've started a study where we're looking at resistance to hydrogen peroxide.

317
00:34:25,870 --> 00:34:35,170
We've been able to select individual clonal lines of sulphur little bits that are resistant to seven 75 micro molar hydrogen peroxide,

318
00:34:35,170 --> 00:34:38,260
a concentration of which the wild type cells cannot grow.

319
00:34:38,260 --> 00:34:43,930
But in blue, we see growth of our resistant lines in presence or absence of hydrogen peroxide.

320
00:34:43,930 --> 00:34:49,440
Essentially, this drug effect off entirely. We sequence them.

321
00:34:49,440 --> 00:34:54,960
Twenty one of these clinical isolates and discovered at our laboratory and had a number of

322
00:34:54,960 --> 00:35:00,480
snips when compared to the reference genome and in many of the things that we sequenced,

323
00:35:00,480 --> 00:35:06,570
we saw deletions, the strange with a large deletion within them and also additional polymorphisms,

324
00:35:06,570 --> 00:35:11,780
which when we analysed the genes, made sense in light of resistance to oxidative damage.

325
00:35:11,780 --> 00:35:20,750
But really exciting to us was the two lines that I've highlighted in blue because these are genetically identical to our wild type lobstering.

326
00:35:20,750 --> 00:35:26,330
There is not a single polymorphism. There are no rearrangements to carry that both Illumina and nanopore sequencing.

327
00:35:26,330 --> 00:35:35,090
They are genetically identical to the wild type and yet are entirely resistant to these concentrations of hydrogen peroxide.

328
00:35:35,090 --> 00:35:39,380
So we wondered obviously what you did as to what was going on.

329
00:35:39,380 --> 00:35:47,270
We profiled the resistant line grown in the absence of hydrogen peroxide that are any second them and

330
00:35:47,270 --> 00:35:53,300
discovered that these lines constructively express the oxidative response genes at a very high level.

331
00:35:53,300 --> 00:35:58,770
So the preconditioned, if you like, for exposure to hydrogen peroxide.

332
00:35:58,770 --> 00:36:06,300
Yeah. Or genetically identical to the wild type. When we come, we wondered if there was some difference in chromosome confirmation.

333
00:36:06,300 --> 00:36:12,300
And when we compared the wild type and this resistant strain, genetically identical resistance genes,

334
00:36:12,300 --> 00:36:20,690
we can note in particular, some quite significant changes in short range context that this look is here.

335
00:36:20,690 --> 00:36:26,300
And this Lucas turns out to correspond to the DPS units, the self Louis counterpart Fahrettin,

336
00:36:26,300 --> 00:36:29,750
so it's involved in modulating iron levels in response stocks with damage,

337
00:36:29,750 --> 00:36:35,270
and this is the principle gene which is upregulated in our resistance lines.

338
00:36:35,270 --> 00:36:39,740
These resistant lines have chromosome conformation alterations,

339
00:36:39,740 --> 00:36:47,120
and so we wondered if that could be put slowing the the determination of resistance phenotype even in the absence of any mutation.

340
00:36:47,120 --> 00:36:53,060
And just to emphasise that this resistance to hydrogen peroxide is highly persistent in the

341
00:36:53,060 --> 00:36:58,190
absence of any hydrogen peroxide treatment and then retesting whether or not they are resistant,

342
00:36:58,190 --> 00:37:06,200
we can take cells out for over 100 generations before we begin to see loss of the phenotype, as we believe that it is a highly persistent phenomenon.

343
00:37:06,200 --> 00:37:16,590
Five minutes great. So we then profiled the cells by 3C chromosome conformation capture as they went through this regression process,

344
00:37:16,590 --> 00:37:22,320
and we're focussing in the box here again. And you can see, you know, the loss of these contacts is different.

345
00:37:22,320 --> 00:37:30,330
Snaps or sourcing blue is no loss of the contact present in the resistant line as we go through the resistance phenotype.

346
00:37:30,330 --> 00:37:41,220
That loss of contact correlates very nicely with a loss of transcription of that little DPS regulation that we see.

347
00:37:41,220 --> 00:37:44,490
So at this point, we thought, well, we're seeing changes in transcription,

348
00:37:44,490 --> 00:37:49,530
we're seeing changes in chromosome confirmation is simply going to be that once again coalescence coming in,

349
00:37:49,530 --> 00:37:56,910
and it's helping to knock down the genes because every other system we've looked at coalescence seem to be acting as a code oppressor.

350
00:37:56,910 --> 00:38:01,020
However, and to a considerable extent, oh, sorry and emphasise,

351
00:38:01,020 --> 00:38:07,290
this is a very specific transcriptional effect that we see this reversion of the phenotype.

352
00:38:07,290 --> 00:38:13,560
We were expecting to see an acquisition of coalescence as we saw this dropping down of transcription.

353
00:38:13,560 --> 00:38:17,380
What we saw was quite the opposite. In the Resistente line,

354
00:38:17,380 --> 00:38:22,240
what we're seeing is that coalition is actually occupying the rebel side and gradually

355
00:38:22,240 --> 00:38:27,850
that peak of abundance drops as we revert back to the wild type sensitive phenotype.

356
00:38:27,850 --> 00:38:35,890
And again, in the absence of any genetic alteration. So we think that cool thing is actually acting essentially to set up an epigenetic

357
00:38:35,890 --> 00:38:41,020
cue in these cells is creating a loop that somehow permissive on a loop structure.

358
00:38:41,020 --> 00:38:46,960
This permissive to the high levels of expression of this does these genes and maintaining

359
00:38:46,960 --> 00:38:51,790
that generation after generation after generation and only grabbing gradually ebbing off.

360
00:38:51,790 --> 00:38:57,040
So we think and the operating hypothesis at the moment, there's obviously some experiments we still need to do.

361
00:38:57,040 --> 00:39:02,440
Is it coalescence functioning to impose this permissive structure that facilitates transcription?

362
00:39:02,440 --> 00:39:12,400
So it's not as simple as being a global coder. Oppressor is actually being used for certain specific tasks to promote transcription as well.

363
00:39:12,400 --> 00:39:21,000
So with that, I'll just finish by thanking the folks that have done the work and the quality of the work was initiated and all the chips.

364
00:39:21,000 --> 00:39:22,900
And I've shown you perform a ritual.

365
00:39:22,900 --> 00:39:31,120
Samson, who now has an independent position at Indiana University and it's a matter who's an independent position back in Kyoto,

366
00:39:31,120 --> 00:39:35,680
now establish the three and high pipe pipelines from the lab.

367
00:39:35,680 --> 00:39:41,290
Peter Ball Hall is doing the epigenetically work with hydrogen peroxide resistance.

368
00:39:41,290 --> 00:39:42,970
The crystallisation was done actually.

369
00:39:42,970 --> 00:39:49,060
My side project I was being department chair wanted to do something, so I made lots of protein and gave them to Giovanni,

370
00:39:49,060 --> 00:39:53,200
who's crystallised and solve the structures for me, which was very nice of them.

371
00:39:53,200 --> 00:39:56,860
But most of all, Dave, you're not on the slide, but thank you again for everything,

372
00:39:56,860 --> 00:40:00,820
for the incredible lectures all the way through your mentorship, your enthusiasm.

373
00:40:00,820 --> 00:40:10,240
You truly have been an inspiration. Thank you. Thank you very much, Steve, that was excellent.

374
00:40:10,240 --> 00:40:18,480
Any questions? Fantastic bacteria of epigenetics say archaea and likely to step outside and discuss that they're all the same thing.

375
00:40:18,480 --> 00:40:24,030
So do you think the formation of this new boundary that gives you the expression of just as the casket process?

376
00:40:24,030 --> 00:40:29,310
It's often and because you're operating on the selection, then they just get maintained.

377
00:40:29,310 --> 00:40:31,740
That was ventilated, but yet we simply don't know at this point.

378
00:40:31,740 --> 00:40:36,030
And that takes us back to the question of how is coalition getting on in the first place?

379
00:40:36,030 --> 00:40:43,560
Something we really want to know, what's looking at, what's unloading it. A million questions that we've asked for a long time and an uncertainty SMC.

380
00:40:43,560 --> 00:40:48,840
We feel that we simply do have answers to.

381
00:40:48,840 --> 00:40:59,280
Fantastic work you had on this slide that there's cohesion after replication, you're also social of always when it divides compacts DNA.

382
00:40:59,280 --> 00:41:05,130
You must have looked at these effects. So we've synchronised sales and performance performed.

383
00:41:05,130 --> 00:41:09,930
Are you asking if you've done chromosome confirmation across the cell cycle? Is that so?

384
00:41:09,930 --> 00:41:16,740
Yeah, we've we've profiled coolest non chromosome confirmation and synchronise cells, and it remains almost unaltered.

385
00:41:16,740 --> 00:41:22,560
We see changes that the power look is so we have power like protein and we do see some conformational changes in those.

386
00:41:22,560 --> 00:41:28,050
And Rachel, that's her power of her independent lab, those working on that side of stuff.

387
00:41:28,050 --> 00:41:35,580
Transcription seems to be the primary determinant for the overall structure. And we've done cell cycle transcriptomics to synchronise cells,

388
00:41:35,580 --> 00:41:40,500
and actually only about 40 open reading frames show significant modulation across the cell cycle.

389
00:41:40,500 --> 00:41:45,450
So, yes, set boundaries are moving, but the compartmentalisation procedure really doesn't alter,

390
00:41:45,450 --> 00:41:51,960
and the fact that we have loci being held together allows us is lovely because although we're dominated by the G2 period of the cell cycle,

391
00:41:51,960 --> 00:41:58,350
while those loci are compacted anyway, so it makes the the technique appropriate for our organism.

392
00:41:58,350 --> 00:42:06,900
Steve, lovely talk that's following up on John's question What what do you know about the nature of cohesion during G2?

393
00:42:06,900 --> 00:42:14,130
Oh, we can sort of second very quick question is your your high C maps?

394
00:42:14,130 --> 00:42:18,400
To what extent are they dependent on coal acid? Mm-Hmm.

395
00:42:18,400 --> 00:42:28,200
Two. OK. So the cohesion that we see as a by fish at multiple loci around the chromosome and just the number of foci.

396
00:42:28,200 --> 00:42:31,950
And the reason we are permitted to do that in the first place prompted to do that in the first place was

397
00:42:31,950 --> 00:42:39,180
that we had observed in two gels the persistence of Hemi Catan injunctions between Dr Lucite and G2 cells.

398
00:42:39,180 --> 00:42:45,060
So we believe that the Hemi cats are probably the causative agent for the cohesion that we see so coalescence in essential gene,

399
00:42:45,060 --> 00:42:53,460
but we cannot express it. And more recently, we've been able to conditionally express mutants and we see if we overexpressed coolies and

400
00:42:53,460 --> 00:42:59,280
then we enhance interactions within the B compartment and intensify the B compartmentalisation.

401
00:42:59,280 --> 00:43:06,660
And if we express the mutant versions of the cells, we begin to see a loss of the compartmentalise identity when we take cells

402
00:43:06,660 --> 00:43:13,620
into stationary phase coalesce and redistributes that out in the chromosome. And at that point, compartmentalisation is lost.

403
00:43:13,620 --> 00:43:19,970
Definitely our last question. OK, so what do you think about everybody standing orders, you know, coalescing,

404
00:43:19,970 --> 00:43:32,160
you identify the international partners or is condensing related or interaction partners that we have are all there's a Tripoli plus protein,

405
00:43:32,160 --> 00:43:36,240
which seems to be a genuine interaction partner.

406
00:43:36,240 --> 00:43:40,700
The rest are all annotated as conserved, hypothetical and even alpha foods are not used to us.

407
00:43:40,700 --> 00:43:49,110
Essentially, nothing like Skippy could be monkey yarmulkes. As to the origin of it, I think it's probably come in an actual chromosomal element.

408
00:43:49,110 --> 00:43:56,310
I think it's probably a bit like is got its own right 50. We've had this drug, 50 like protein come in and it's been co-opted into functioning.

409
00:43:56,310 --> 00:43:59,850
Even its location in the chromosome is near one of the origins of replication.

410
00:43:59,850 --> 00:44:05,880
That itself was a relatively recent evolutionary acquisition, at least from the lineage tracing that we've done.

411
00:44:05,880 --> 00:44:10,260
So I mean, why the progenitors of the clinical lost condensed in the first place?

412
00:44:10,260 --> 00:44:13,650
I have no idea I'd love. I'd love to hear people's thoughts in that,

413
00:44:13,650 --> 00:44:20,070
but it looks as though different lineages have co-opted different proteins to fulfil some sort of structuring role.

414
00:44:20,070 --> 00:44:22,080
And again, are we seeing a truly condensed and late?

415
00:44:22,080 --> 00:44:30,600
I do think we are, but we're seeing, you know, at least the level of structuring being imposed by this superfamily protein.

416
00:44:30,600 --> 00:44:34,280
Thank you very much. OK, thanks. Thanks.

417
00:44:34,280 --> 00:44:44,390
So, first of all, thanks to Dave and Lydia for getting me out of my house first time across the ocean in two years, I wouldn't miss this for.

418
00:44:44,390 --> 00:44:46,070
For anything, right, so thank you.

419
00:44:46,070 --> 00:44:56,690
And of course, this is a shout out to Lydia, who you may recognise her last name has some vague genomic similarity in terms of code on usage to mine.

420
00:44:56,690 --> 00:45:03,920
And so when she asked me for a title, I said, Of course, using your motors, twist the mouse, right?

421
00:45:03,920 --> 00:45:09,950
So which loosely translated is, if you can't twist, then you got it right.

422
00:45:09,950 --> 00:45:14,600
So, so that's the title of my talk. It actually has nothing to do what I'm going to say.

423
00:45:14,600 --> 00:45:19,880
Rather, it's a it's a it's a statement actually on life, I realised, because, you know,

424
00:45:19,880 --> 00:45:24,470
if you think about it, there are lots of times you need to twist out of something.

425
00:45:24,470 --> 00:45:29,830
You know, the twist a little bit, and if you don't, you're going to rise with some kind of pain.

426
00:45:29,830 --> 00:45:34,550
And so it was coming home late from the pub or, you know, explaining something else.

427
00:45:34,550 --> 00:45:41,450
So this this is actually a very broad statement on on life that I could probably stop right now.

428
00:45:41,450 --> 00:45:48,860
So but you know, since really there's no connexion academic, at least between Dave and I.

429
00:45:48,860 --> 00:45:58,850
So I'm here because I can give a completely fabricated account of his accomplished accomplishments and contributions to science and which I will.

430
00:45:58,850 --> 00:46:12,210
So, so let me, first of all, tell you what inspired me to have this title, which, as I said, is only vaguely related to what I'll talk about.

431
00:46:12,210 --> 00:46:18,110
You know, it turns out that Dave, actually, when he was doing preps I learnt was doing would always sing a song.

432
00:46:18,110 --> 00:46:21,890
He had a song he just loved, and it was a twist in right.

433
00:46:21,890 --> 00:46:27,200
So if those of you were wondering what that song is, you know, he adapted.

434
00:46:27,200 --> 00:46:32,150
Of course, there a a group called The Beatles, I think some of you were actually familiar with them.

435
00:46:32,150 --> 00:46:38,480
And they did a song. It was actually cover, not their original of a twist and shout, but they've had this.

436
00:46:38,480 --> 00:46:44,090
This has this. And he's the only guy I know who's called his dad a baby. So and he was doing a prep.

437
00:46:44,090 --> 00:46:47,990
These licence sales. So shake it up, baby. Now twist and ride, you know?

438
00:46:47,990 --> 00:46:54,140
Come on. Come on, baby. This is a day prep. I won't go through all the steps, you know, in the middle, in the middle here, he says.

439
00:46:54,140 --> 00:46:57,410
You know, you twist so fine. You know, this isn't your little DNA.

440
00:46:57,410 --> 00:47:02,180
And you know, actually, unbeknown to Dave, at dinner today, he's going to sing this song.

441
00:47:02,180 --> 00:47:08,270
So, um, and then, you know, let me know that you're mine, all this stuff.

442
00:47:08,270 --> 00:47:13,790
And then, of course, you know, as he gets down to the end of the prep, he does final extraction.

443
00:47:13,790 --> 00:47:21,590
Well, shake it, shake it, shake it, shake shake a bit. So, so so this this is a well-kept secret, you know of Dave's.

444
00:47:21,590 --> 00:47:26,540
She never let on to anybody in his lab that that he was doing it this way.

445
00:47:26,540 --> 00:47:32,080
So and then my first meeting with Dave was actually at Nesi Bridge.

446
00:47:32,080 --> 00:47:38,860
Yeah, I'll send a copy to anybody who wants this. And but especially if Dave agrees to singing it, so.

447
00:47:38,860 --> 00:47:44,990
So I met Dave. My recollection was that nephew bridge and I managed to get the same video that Lorraine got.

448
00:47:44,990 --> 00:47:50,090
I just have a different take on this, this this episode.

449
00:47:50,090 --> 00:47:58,210
So our conversation, at least my questions, is something like, OK, Dave, what's a haggis?

450
00:47:58,210 --> 00:48:04,910
You know? And then of course, you saw this. You'll see it again. Now, I don't have the sound on, but it's probably just as good sound.

451
00:48:04,910 --> 00:48:10,760
It has the same impact. And I said, Oh, OK, you throw it.

452
00:48:10,760 --> 00:48:14,330
And Dave proceeded to do this. This wonderful throw.

453
00:48:14,330 --> 00:48:19,850
Right? Yes, it explodes. Is something I have. I still have no idea.

454
00:48:19,850 --> 00:48:26,210
And so and then of course, I said, Oh, and then you eat it.

455
00:48:26,210 --> 00:48:29,960
And you know, at that point, I said to Dave, OK, I got it.

456
00:48:29,960 --> 00:48:36,800
I'll have a whisky. So. So that that was my experience with Dave at least meeting him.

457
00:48:36,800 --> 00:48:41,660
And I have to say, over the years, I've had a chance to see him at other events.

458
00:48:41,660 --> 00:48:47,030
And you know, here's here's here he is in CIAC am fortunate I don't have too many photos.

459
00:48:47,030 --> 00:48:50,690
So fortunately, I have a few rod contribute. I think this one. There he is, talking to one.

460
00:48:50,690 --> 00:48:55,880
Alonso in the next one is the reaction from Mike Cox when he's explaining one of his great theories.

461
00:48:55,880 --> 00:49:00,650
And Mike is like just sucking in, sort of holding it back right there so that, you know,

462
00:49:00,650 --> 00:49:05,870
these are fractured fairy tales like I could put in whatever little word bubbles I want.

463
00:49:05,870 --> 00:49:10,970
And here he is. Actually, he's he's he's making a pact with the devil.

464
00:49:10,970 --> 00:49:21,480
This is Lucifer right here. You can see the Yeah. And look, Lorraine is witnessing witnessing this pact so and so you know,

465
00:49:21,480 --> 00:49:27,320
if you see a certain group of people hanging around and you know, you know, it's sad, but this, I have to put it in here.

466
00:49:27,320 --> 00:49:30,920
This is the only slide of half of Dave actually attending a lecture. So he did.

467
00:49:30,920 --> 00:49:35,600
He did attend lectures. He wasn't just going off eating, drinking, sightseeing.

468
00:49:35,600 --> 00:49:43,850
OK, so. All right. So now now more fractured fairy tales, more.

469
00:49:43,850 --> 00:49:49,580
This is what I'm going to talk about. You can take this with a tongue in cheek. This is, I hope, hopefully it's not not the fiction.

470
00:49:49,580 --> 00:49:54,800
So we've been interested in double holiday junctions and topology.

471
00:49:54,800 --> 00:50:03,300
And what I'm going to tell you is that the resolution of double holiday junctions is actually complicated and it's affected by motor protein,

472
00:50:03,300 --> 00:50:11,450
so for some reason and nucleases. So this work done by Katsumi more Myanmar too, and also Jodi Plank for In My Lab.

473
00:50:11,450 --> 00:50:14,900
And now here are simple questions and simple and straightforward answers.

474
00:50:14,900 --> 00:50:18,410
So is a double holiday junction simply to a single holiday?

475
00:50:18,410 --> 00:50:28,190
I think many of you know, but to make clear, no, it's not just to single holiday options and can top equal equally top of one tree and wreck.

476
00:50:28,190 --> 00:50:31,970
You dissolve double hardly junctions. And again, I'll tell you why I asked this question.

477
00:50:31,970 --> 00:50:39,230
This simple answer is yes. And then the third question is, can a double holiday junction be resolved politically?

478
00:50:39,230 --> 00:50:48,390
And actually, unless unless I've missed it in literature, no one has ever tested whether a double holiday junction can be cleaved.

479
00:50:48,390 --> 00:50:51,630
This is important. It's in every war textbook, every textbook. Right?

480
00:50:51,630 --> 00:50:57,390
And the fact is there is not again, unless I missed it, there's no piece of data out there.

481
00:50:57,390 --> 00:51:01,950
And the answer is yes, it can. But the outcome is not as simple as you might have expected.

482
00:51:01,950 --> 00:51:09,960
All right. So here's the model from Robyn Holiday in 1960 for a prescient model, the holiday,

483
00:51:09,960 --> 00:51:17,370
the structure in the middle that bears his name now explains many of the phenomena of recombination,

484
00:51:17,370 --> 00:51:22,770
including including crossing over and gene conversion.

485
00:51:22,770 --> 00:51:26,760
So I won't go over that. Many you know this even better than I do so.

486
00:51:26,760 --> 00:51:34,500
So that's a holiday junction and he proposed cleavage and this is you get this or that depending on whether you have horizontal vertical cleavage,

487
00:51:34,500 --> 00:51:39,780
right? And now here, fast forward to nineteen eighty three.

488
00:51:39,780 --> 00:51:43,800
This is a simple rendition of the double strand break repair model.

489
00:51:43,800 --> 00:51:51,900
And when it was appreciated that the sites of initiation also replication, stalling and cleavage replication sites leads to a double strand breaks.

490
00:51:51,900 --> 00:51:55,260
Lorraine actually went through some of the processing biochemical steps.

491
00:51:55,260 --> 00:51:59,340
Here you get to holiday junctions. There they are, right?

492
00:51:59,340 --> 00:52:09,510
And then again, what is very clear and very accurately stated this is from the from the paper resolution of two junctions by cleaving either

493
00:52:09,510 --> 00:52:17,010
inner or outer strands leads to two possible crossover and two possible to balance across the crossover configurations.

494
00:52:17,010 --> 00:52:21,150
And this is absolutely correct. OK. So I'm not contesting this one bit.

495
00:52:21,150 --> 00:52:28,860
But when you look at this picture and you teach it in class, you would get the impression that you're always cleaving double holiday junctions.

496
00:52:28,860 --> 00:52:35,220
That's that's historical. But they were trying to explain the existence of crossovers.

497
00:52:35,220 --> 00:52:43,470
So it turns out the reason why, I said, are two holiday junctions equivalent to to a double holiday junction answer is no,

498
00:52:43,470 --> 00:52:50,430
because there is another pathway for this taking care of removing these holiday junctions,

499
00:52:50,430 --> 00:52:55,740
which would cause an endless junction chromosome segregation problems, et cetera,

500
00:52:55,740 --> 00:53:04,420
is and that turns out it's a family of proteins, a helicase and took out type three type II sunrays.

501
00:53:04,420 --> 00:53:05,370
And again, it's important for you,

502
00:53:05,370 --> 00:53:12,840
those of you who don't remember you're told by some races that Topo three from basically all organisms is a type one type II summaries.

503
00:53:12,840 --> 00:53:20,580
And what that means is it passes. It cleaves in one strand and passes once that cleave strand once cleaved strand at a time

504
00:53:20,580 --> 00:53:26,760
where type 2s make a double strand break and past past DNA through the double strand break.

505
00:53:26,760 --> 00:53:31,440
So this is an AFM image. Frank Hartman took my lab a long time ago,

506
00:53:31,440 --> 00:53:42,780
showing that Recue and Toeplitz where you can actually cannot indicate duplex DNA and this must occur through concerted single Strand Passage events.

507
00:53:42,780 --> 00:53:46,860
And I also put the word concerted in because we've never detected a hem I can't make.

508
00:53:46,860 --> 00:53:55,690
They're always full cat names. Now, again, going back to the serious part of science, I've had extensive discussions with Dave.

509
00:53:55,690 --> 00:53:56,680
Nick Khazali,

510
00:53:56,680 --> 00:54:08,230
Ian Hicks and I remember talking about how can you test a strand and get against and how do you test for passage of a double holiday junction?

511
00:54:08,230 --> 00:54:12,580
And there really is no easy way to make a double holiday junction.

512
00:54:12,580 --> 00:54:17,950
And the first one who succeeded to make something that looks like the holiday junction again is Ian Hickson.

513
00:54:17,950 --> 00:54:26,080
There's a paper he published. These are some oligo substrates, and it's I'm being facetious and also polite.

514
00:54:26,080 --> 00:54:34,060
This is this is the first attempt to make a double holiday junction, but the call is short is really being kind.

515
00:54:34,060 --> 00:54:38,260
It has a linking number of only about under two. I think it's one point sixty five, they estimated.

516
00:54:38,260 --> 00:54:43,900
So it's barely has any linkage. I mean, it is topological at length.

517
00:54:43,900 --> 00:54:51,400
And what he showed for the Bloom's protein that these proteins can dissolve the this this structure,

518
00:54:51,400 --> 00:54:55,150
which is, you know, I will accept the fact it's a double holiday junction.

519
00:54:55,150 --> 00:55:02,440
But we all said it'd be nice if somebody can do something bigger, a bigger DNA substrate with a larger linking number.

520
00:55:02,440 --> 00:55:06,910
And this came along when Jodie Planck was working Cowshed Duke.

521
00:55:06,910 --> 00:55:13,180
The forecourts passed away after significant effort.

522
00:55:13,180 --> 00:55:17,140
Jodie made the substrate of by begin a process.

523
00:55:17,140 --> 00:55:23,470
I won't show you this. There's a dozen steps and this the blue and red are non homologous regions.

524
00:55:23,470 --> 00:55:29,080
The black are regions that are identical in sequence, and there's a holiday junction at each end.

525
00:55:29,080 --> 00:55:35,950
And again, this is a DNA with no twist. If you try to picture it with twist, that's what it's going to look like.

526
00:55:35,950 --> 00:55:42,130
So you cannot separate these by just pulling them. They're interlinked. And in fact, this this one is.

527
00:55:42,130 --> 00:55:48,700
So do you still remember your topology? This is all twists and almost no ride if you accept the fact it's lying flat in that plane.

528
00:55:48,700 --> 00:55:52,090
So but what we wanted to know is, how do you I mean,

529
00:55:52,090 --> 00:55:57,550
you have cartoons that show this thing branch migrating a holiday junction principle can branch migrate?

530
00:55:57,550 --> 00:56:05,350
But it's not that simple if it's top logically linked. And again, what we're able to do and I'll show you the data for E. coli Petrushka.

531
00:56:05,350 --> 00:56:10,600
When I say that we Peter, when he was a postdoc in my lab, showed that you can dissolve.

532
00:56:10,600 --> 00:56:15,100
So the reason to holiday junctions or double holiday chunks is not just the

533
00:56:15,100 --> 00:56:19,090
sum of two single holiday chunks is without going through the detailed steps,

534
00:56:19,090 --> 00:56:24,910
you can take any either this junction or that junction and they can cause and converge them.

535
00:56:24,910 --> 00:56:31,450
One can move or both can move towards each other. And here you wind up with a penultimate separated product.

536
00:56:31,450 --> 00:56:38,980
This is a singly linked Hemi cattani. And then as long as you and you remember you're passing through one strand at a time,

537
00:56:38,980 --> 00:56:42,850
you have to pass the parental strands and then you can separate them totally.

538
00:56:42,850 --> 00:56:47,440
And the important thing about the whole process here from the biology is not a crossover.

539
00:56:47,440 --> 00:56:49,330
You never get a crossover, right?

540
00:56:49,330 --> 00:56:55,930
And again, if you think about recombination, I mean, the history of recombination and I'm an outsider, my degrees are in chemistry.

541
00:56:55,930 --> 00:56:57,670
So you know what?

542
00:56:57,670 --> 00:57:05,580
They all were designed to largely explain crossovers, not because crossovers were the most important thing in the world, but that's what you ask for.

543
00:57:05,580 --> 00:57:11,620
Now there are other things, of course, that we're being asked for, but the models had to explain crossovers mismatch, all this other stuff.

544
00:57:11,620 --> 00:57:15,940
But the reality is, no chromosome really wants to crossover.

545
00:57:15,940 --> 00:57:20,530
If you cross over a circular E coli chromosome, you get a dimer and segregation problem,

546
00:57:20,530 --> 00:57:26,620
and we're going to hear that there are specific mechanism to if in fact worked on these proteins to get rid of those dimers.

547
00:57:26,620 --> 00:57:31,420
And then on linear chromosomes, you get translocations, right, so you don't want to crossover.

548
00:57:31,420 --> 00:57:39,400
And this is one mechanism for non non crossover. And in fact, its predominant mechanism in eukaryotes.

549
00:57:39,400 --> 00:57:46,570
All right. So now I'm going to I worked on so work and Nicole, I actually worked on T for phage therapy for in the old days,

550
00:57:46,570 --> 00:57:52,060
and I wanted to ask whether you call a total of three and you dissolve double holiday junctions.

551
00:57:52,060 --> 00:58:00,580
And here's a list that showed some similarities of the RECUEILLIR cases and type one a two-part summarises.

552
00:58:00,580 --> 00:58:08,440
And why would I ask this question? Well, in the eukaryotes, both segments obviously a one and human bloom.

553
00:58:08,440 --> 00:58:14,590
There's a domain which is called TR for the Subway Series three, my one two interaction domain.

554
00:58:14,590 --> 00:58:20,710
I want to tell you about one or two of the subunits basically of the dissolving zone, if you want to call it that.

555
00:58:20,710 --> 00:58:28,990
But Rick E. Coli Recue has no TR domain, and the one thing I should also add is we we attempted to use every, every other.

556
00:58:28,990 --> 00:58:37,630
One of the human helicase is to substitute for Bloom and none of them substitute in the the dissolve dissolution reaction.

557
00:58:37,630 --> 00:58:44,120
So this appeared to be essential, important for for dissolution.

558
00:58:44,120 --> 00:58:50,060
So at this point, cuts would be more inmates who have purified all these proteins did the reaction.

559
00:58:50,060 --> 00:58:54,830
And again, let me just I'm not going to go through a lot of data.

560
00:58:54,830 --> 00:58:57,910
I just want to point out you start out with these substrates, the substrate,

561
00:58:57,910 --> 00:59:03,230
the double holiday junction substrate and through branch migration and strand passage,

562
00:59:03,230 --> 00:59:15,600
you separate them into two basically monomer circles a piece that might not make OK, but I know everyone leave, but I thought that was a miracle.

563
00:59:15,600 --> 00:59:21,620
And whereas if you had a crossover, you get this dimer and a b dimer.

564
00:59:21,620 --> 00:59:26,030
So it's very clear when you get a non crossover versus a crossover.

565
00:59:26,030 --> 00:59:32,210
And if you have all three proteins, top of three recue as this B, this is the optimal reaction.

566
00:59:32,210 --> 00:59:38,720
You get the product a the Monroe Circle product b, the other monomer circle and no crossover.

567
00:59:38,720 --> 00:59:44,270
The crossovers would run above the holiday junction substrate. So zero zero crossover.

568
00:59:44,270 --> 00:59:45,050
But the remarkable,

569
00:59:45,050 --> 00:59:55,910
not remarkable pleasant thing for us was that this protein from E. coli with no top brass hammers three domain still did the solution.

570
00:59:55,910 --> 01:00:01,700
So this is just to show you this this the same product, same behaviour with the eukaryotic S. Just one protein.

571
01:00:01,700 --> 01:00:11,550
So we never get crossovers. And so E! Coli has a mechanism to facilitate and in fact, predominantly encourage non crossovers.

572
01:00:11,550 --> 01:00:18,500
All right. So all the data I'm not going to show you if this solution is very efficient, it only takes one animal or of each of these proteins.

573
01:00:18,500 --> 01:00:28,250
It doesn't require a cognate helicase, so it is promiscuous. We can substitute this one and we can even rob which I'll come back to in a little bit.

574
01:00:28,250 --> 01:00:37,160
It does not require a cognate as B you can substitute, but E. coli Topa one a different type on top of this cannot.

575
01:00:37,160 --> 01:00:45,710
And that's because there's a there's a decapitation loop in the type three proteins or in the Army one codon subunits,

576
01:00:45,710 --> 01:00:51,290
and there is no physical interaction to rescue until three, in contrast again to the UK proteins.

577
01:00:51,290 --> 01:00:56,270
But importantly, only C-terminal stresses. But importantly, this solution is concerned.

578
01:00:56,270 --> 01:01:03,500
So you need simultaneous action on top of three and recue. They have to be working on that DNA molecule or put more accurate scientifically.

579
01:01:03,500 --> 01:01:07,370
We put them both in the same two with the DNA. They both be prosperous the same time.

580
01:01:07,370 --> 01:01:12,170
You can't put them in sequentially, so it's a concerted reaction.

581
01:01:12,170 --> 01:01:15,410
Actually, the you can think about the mechanism is not really known.

582
01:01:15,410 --> 01:01:24,920
I mean, the the structures of these individual proteins are available, but but it's not really clear how they're acting together on the junctions.

583
01:01:24,920 --> 01:01:29,810
OK, so now can a double holiday junction be resolved politically?

584
01:01:29,810 --> 01:01:33,680
And you know, I say, why would you even ask this question? Because it's in every textbook.

585
01:01:33,680 --> 01:01:37,820
We know that it's in a textbook and must be right. OK, so not.

586
01:01:37,820 --> 01:01:42,470
Not really. And so. Well, you know, I told you they're interlinked cat named.

587
01:01:42,470 --> 01:01:45,560
So first of all, you don't need to believe them.

588
01:01:45,560 --> 01:01:52,250
I mean, you do need to cleave them if you want explain crossovers because I just told you this pathway does not give you crossovers.

589
01:01:52,250 --> 01:01:59,450
And then structurally double holiday searches are composed of two parallel structures and most cleaving enzymes.

590
01:01:59,450 --> 01:02:07,880
For example, Rossi, this they lillia. Others have shown that to bind to a preferred open a. parallel structure and

591
01:02:07,880 --> 01:02:12,260
single how the junctions can adopt both a parallel and an interior configuration,

592
01:02:12,260 --> 01:02:15,830
but double holiday junctions are constrained to a parallel configuration.

593
01:02:15,830 --> 01:02:21,170
And then finally, the persistence, like the DNA is about fifty three nanometre, which 150 base pairs.

594
01:02:21,170 --> 01:02:26,630
So unless the intervening being more than two persistance like the holiday junctions connected by rigid rods,

595
01:02:26,630 --> 01:02:34,070
so the local deformation to put them into an open conformation is the energetic start to become more and more difficult.

596
01:02:34,070 --> 01:02:40,550
So basically, Katsumi took the double holiday junction, remember, you can get two products,

597
01:02:40,550 --> 01:02:48,800
you can get a non crossover crossover, very diagnostic dimer circle or Dimer Circle or Monroe Circles.

598
01:02:48,800 --> 01:02:56,000
And what he did when he threw in Rosie, he showed basically he was mostly non crossover.

599
01:02:56,000 --> 01:03:02,870
And I show you a concentration dependent because I'm not going to I'm not going to torture you through the conflict to get to the conclusion.

600
01:03:02,870 --> 01:03:08,510
We actually, I'll tell you. I mean, they were stupid, but like we kind of were saying, huh?

601
01:03:08,510 --> 01:03:14,270
At first, first thing was, OK, there's a bias. It was structural. I just told you the structure of the holodeck.

602
01:03:14,270 --> 01:03:21,560
They junction is different than the single holiday junction. But when I did a concentration dependent, we saw that at low concentrations,

603
01:03:21,560 --> 01:03:26,120
you predominantly got non-profit crossovers by a large margin over crossovers.

604
01:03:26,120 --> 01:03:33,740
And only when you added more and more reps say did you start to equilibrate or get to a point where both are plateauing and if shown as a ratio,

605
01:03:33,740 --> 01:03:40,130
this is maybe an easier plot to see as you put in. You get to about 30 some odd percent.

606
01:03:40,130 --> 01:03:45,200
That's the maximum. But there's a concentration of parts and the first the first thing.

607
01:03:45,200 --> 01:03:50,840
But I think we were we thought about why should there be a concentration dependence in this crossover?

608
01:03:50,840 --> 01:03:56,690
Non crossover distribution? Five minutes. OK, good. Thank you. And so we said, Oh, you know, kind of being done first.

609
01:03:56,690 --> 01:03:58,010
We don't have a row of a b.

610
01:03:58,010 --> 01:04:05,300
The motor proteins that are partners of great worth by Steve West many years ago show that they accelerate roughly cleavage.

611
01:04:05,300 --> 01:04:11,570
So we threw in and here's what they look like a simple structure, cartoon structure.

612
01:04:11,570 --> 01:04:18,890
And so we we threw it in. And here's the result. So instead of getting none.

613
01:04:18,890 --> 01:04:23,630
Instead of getting more crossovers, we went down to almost zero.

614
01:04:23,630 --> 01:04:29,750
So the more rough C of a b you put in. The less crossover you get.

615
01:04:29,750 --> 01:04:39,140
So basically, you make the rough sea protein seem to have a rough, a dependent bias on crossover non crossover distribution.

616
01:04:39,140 --> 01:04:42,770
And again, we scratch their heads and said, what the [INAUDIBLE] is going on?

617
01:04:42,770 --> 01:04:46,250
Because we came to it from a structural consideration,

618
01:04:46,250 --> 01:04:51,720
we thought structurally a double holiday junction should be different than a single holiday junction.

619
01:04:51,720 --> 01:04:57,230
And it is OK. But that wasn't the answer, right? So I'm going to cut to the chase to five minutes.

620
01:04:57,230 --> 01:05:05,630
There's the data. You're almost down at zero crossovers when you have the highest rob concentration we could put in, that's the motor protein.

621
01:05:05,630 --> 01:05:11,570
And so here's a what's the explanation?

622
01:05:11,570 --> 01:05:15,500
The slow computer here. OK, here we go.

623
01:05:15,500 --> 01:05:19,070
Well, the answer is going to be cut and migrate and I'll tell you in a minute.

624
01:05:19,070 --> 01:05:28,580
So here's the canonical model. So if you have a double holiday junction in circles, you're going to get the cross, the non crossovers and a crossover.

625
01:05:28,580 --> 01:05:34,100
These are all the products you would get depending on whether you got horizontal cleavage or a vertical cleavage did this.

626
01:05:34,100 --> 01:05:46,120
So these are the four products you get. But. It was a no dull moment when we realised that if you just cut one and branch, migrate to the other,

627
01:05:46,120 --> 01:05:52,480
to the other junction, you get all four of the possible non crossover products.

628
01:05:52,480 --> 01:06:01,690
That is all for possible plasmid molecules. So the note no dull moment was I don't know about the rest of you in the audience,

629
01:06:01,690 --> 01:06:05,720
but if you really think about the proposed mechanism, it's a cartoon, right?

630
01:06:05,720 --> 01:06:12,820
That the last model? It's a cartoon. It says yes, if you cleave it both, you will get those products.

631
01:06:12,820 --> 01:06:19,720
Biology in life doesn't work like that. You got to junctions, you got two proteins, unless there's some coordination between them, you cut one.

632
01:06:19,720 --> 01:06:25,300
The other one's just lollygagging around. You cut the other one. And and when do you cut to?

633
01:06:25,300 --> 01:06:29,590
Only if you have high concentrations of a cleaving enzyme, the Rossie,

634
01:06:29,590 --> 01:06:37,960
if you cut one and a motor protein comes along or the reason I'm skipping a lot of data, even we did not have more of a b.

635
01:06:37,960 --> 01:06:41,710
We got more non crossovers. Their spontaneous branch migration.

636
01:06:41,710 --> 01:06:45,580
The reason you can't make double holiday junction substrates easily is when you make them.

637
01:06:45,580 --> 01:06:47,500
They've spun a single one, for sure.

638
01:06:47,500 --> 01:06:51,910
And the doubles, if you don't do it right, they'll branch, migrate fully homologous or branch back right off the ends.

639
01:06:51,910 --> 01:06:57,820
So it was like one of these. I mean, I had a lot of dents in my forehead because all throughout my life, you know,

640
01:06:57,820 --> 01:07:02,860
I just flat myself in the head, you know, like, this is just dumb, you know?

641
01:07:02,860 --> 01:07:10,780
And so, so basically. So the answer is equal, I reckon so it can resolve the functions, not cross over rough.

642
01:07:10,780 --> 01:07:17,620
So you can cut one junction of a double holiday junction and rub can migrate the other holidays towards the other side and not cross over.

643
01:07:17,620 --> 01:07:24,190
And crossovers form after cleavage of one, but only if the second one is cut before each branch migrated off.

644
01:07:24,190 --> 01:07:30,970
Now here's here's the like some 3D three twisty like molecules those of you who like something more accurate.

645
01:07:30,970 --> 01:07:36,820
This is trying to show you dissolution. OK, I to show this cuts the Tatsumi days to make this.

646
01:07:36,820 --> 01:07:42,970
And then, of course, if you cleave on both simultaneously, you get the here's a crossover product.

647
01:07:42,970 --> 01:07:49,180
But you know what you can do, you know painfully in cartoon form, show everyone is if you cut here and branch,

648
01:07:49,180 --> 01:08:00,700
migrate the other junction to the neck, basically non crossover, you cut there and then you have rubber be cut to cut to the side, not cross over.

649
01:08:00,700 --> 01:08:05,860
You could do this with the other cuts I took. So basically, you get you always get non crossovers.

650
01:08:05,860 --> 01:08:13,000
So if you put this in a broader context, of course, in addition to break repair model, when you get down to this point, it's losing power here.

651
01:08:13,000 --> 01:08:20,170
But when you get down to the point of a double holiday junction or a single holidays, a single holiday traditions must be nuclear logically resolved.

652
01:08:20,170 --> 01:08:23,860
They cannot be there. They're not topological molecules. They must be cleaved.

653
01:08:23,860 --> 01:08:29,560
And so here you get the probably expected 50 50 of crossover, not crossover.

654
01:08:29,560 --> 01:08:37,300
But if you go through that double holiday junction and dissolve root, or if you go to a double holiday junction route and nick and migrate,

655
01:08:37,300 --> 01:08:40,690
you always get not processed and a non crossover, not cross or not cross over.

656
01:08:40,690 --> 01:08:48,010
And the only time you get crossover as a minor species is with resolution primarily of a single holiday junction.

657
01:08:48,010 --> 01:08:54,280
But in theory, it's possible to get it with, of course, a double holiday junction. So now you always think things are novel, right?

658
01:08:54,280 --> 01:08:56,440
And I should have said at the beginning, sorry, Kim.

659
01:08:56,440 --> 01:09:02,210
Kim actually proposed an annual review article about topological dissolution of double holiday junctions.

660
01:09:02,210 --> 01:09:08,110
So I forgot to say that. So every time you think you do something novel, you find somewhere, something somebody.

661
01:09:08,110 --> 01:09:12,340
I thought about it earlier. Now I joked last time.

662
01:09:12,340 --> 01:09:18,190
How about Nick, Nick and migrate? It turns out this had been published. And here, here's so here it is.

663
01:09:18,190 --> 01:09:23,800
It's right here. It's very clear to all of you. And if that's not clear, you can look at an alternate one minute.

664
01:09:23,800 --> 01:09:30,610
You can look at the alternative model. And this is makes it very clear to everybody in this audience that it was totally opaque, right?

665
01:09:30,610 --> 01:09:36,820
So frank stuff published in Ninety Nine, basically that you can do this and it's hidden in here.

666
01:09:36,820 --> 01:09:42,100
And this actually reminds me of Nessie Bridge. So I want to just spend one moment on this slide because when you look at this,

667
01:09:42,100 --> 01:09:46,270
it's got all the things that used to give me like nightmares I couldn't understand. I'm a I'm a dump chemist.

668
01:09:46,270 --> 01:09:50,470
You know, the four floors, a sixty two, there's a five two three down here, and I looked around,

669
01:09:50,470 --> 01:09:55,960
I couldn't find the average Ford or Ford's, which I always love because I never understood. But you know, basically,

670
01:09:55,960 --> 01:10:04,360
this is when you find that little part of Frank's cartoon that that is the way I look at the world and you get double holiday junction,

671
01:10:04,360 --> 01:10:14,380
you can dissolve it. He's basically had here the solution, and here he's got a migration of one holiday to a next site.

672
01:10:14,380 --> 01:10:20,500
And in both cases, you get no crossovers. So I must say I didn't see this paper until we dug around.

673
01:10:20,500 --> 01:10:24,820
But so my last slide here is Dave.

674
01:10:24,820 --> 01:10:27,550
Congratulations, thank you for exceptional creative science.

675
01:10:27,550 --> 01:10:34,870
And as part of my fabrication of Dave Dave, here he's showing me around and in Oxford, he's telling me he owns everything.

676
01:10:34,870 --> 01:10:48,190
As far as the eye can see, this is his garden. You know everything. So thank you, Dave, for a good time to.

677
01:10:48,190 --> 01:10:55,780
Thanks to a couple of questions. OK. OK.

678
01:10:55,780 --> 01:11:01,630
So coming back to your title, moving holidays, junctions has twists and dried consequences.

679
01:11:01,630 --> 01:11:08,410
So there's those circular substrates you're using. They are fully top, logically closed with with no nicks.

680
01:11:08,410 --> 01:11:12,990
Yeah, right. That is guaranteed, is it? Yeah, we've checked that that right?

681
01:11:12,990 --> 01:11:20,740
Correct. And remember, they can't even migrate outside of the region, homology to to the right because there's there's no homology there.

682
01:11:20,740 --> 01:11:30,790
So they're they're really bounded on one side. Story with Dave is a little bit different as we'll get to, and I'll explain who I am in a minute,

683
01:11:30,790 --> 01:11:39,700
but I just want to say that's because there's one discovery that everybody has not mentioned about Dave was that in terms of the key question,

684
01:11:39,700 --> 01:11:45,940
they asked about the key components. And he asked it to me when we.

685
01:11:45,940 --> 01:11:51,400
Well, many of the places that we met, which haven't been in laboratories.

686
01:11:51,400 --> 01:11:58,750
And he answered it because the key component for Dave, he answered the question What's the key component in life?

687
01:11:58,750 --> 01:12:03,610
Is Lydia, who is at the front driving. It's Lydia.

688
01:12:03,610 --> 01:12:12,950
So I want to thank Lydia for all the organisation that slows down.

689
01:12:12,950 --> 01:12:18,500
So who am I, I'm actually a physicist and mathematician, don't worry, I'm between you and lunch.

690
01:12:18,500 --> 01:12:22,310
I'm not going to give you a bunch of formulae, but it's another side that day.

691
01:12:22,310 --> 01:12:27,410
But I think, you know everyone here, you've met him in the lab. I met Dave outside a lab.

692
01:12:27,410 --> 01:12:35,660
In fact, it was about 20, three or four years ago, and I'm not kind of sure whether the first was involved standing outside the old field and gym,

693
01:12:35,660 --> 01:12:39,140
or we had something in our hands and we were drinking. It was one of those two.

694
01:12:39,140 --> 01:12:44,990
I like to think it was the first one being kind of good dads, but they were once asked me,

695
01:12:44,990 --> 01:12:48,950
you know, because I was ever in the physics department at the time here in Oxford.

696
01:12:48,950 --> 01:12:52,790
And you know, what do you do? I do this when you do maths.

697
01:12:52,790 --> 01:12:56,820
So what's wrong with maths? Is that what do you mean, what's wrong with that?

698
01:12:56,820 --> 01:13:01,410
There's nothing wrong with maths and physics. We've got big building. You know, we teach.

699
01:13:01,410 --> 01:13:06,240
We have a lot of students. We just develop new techniques. He said, Yeah, but what?

700
01:13:06,240 --> 01:13:13,320
Why is it? We don't see more maths over here. You know, certainly not 20 feet away or 60 feet.

701
01:13:13,320 --> 01:13:22,680
You know what was going on? What? You know why? Is it that just, you know, you don't want to come over, which may be true?

702
01:13:22,680 --> 01:13:31,680
You never know. But it made me think. And since then, I've just been working on this problem of kind of what's wrong with that?

703
01:13:31,680 --> 01:13:35,490
And I want to explain to you why it might be relevant to some of the things that

704
01:13:35,490 --> 01:13:39,960
you were talking about today and why you may not have seen many people over here.

705
01:13:39,960 --> 01:13:46,650
And yet they have their own conferences and they have their own journals and they do their own things and sometimes they ask the web biology in it.

706
01:13:46,650 --> 01:13:53,460
And yet it doesn't make it over here. So I want to run through what I will do it very quickly.

707
01:13:53,460 --> 01:14:00,770
What is the problem with maths? And the amazing thing about maths is it's kind of a jack of all trades master.

708
01:14:00,770 --> 01:14:09,050
You know, you can apply the same, the same maths gets applied to COVID as it does to infection models.

709
01:14:09,050 --> 01:14:15,770
As know, a lot of this was developed in the operations research period of the post-World War One,

710
01:14:15,770 --> 01:14:19,610
when they began to realise they need to put numbers on casualties and things like this,

711
01:14:19,610 --> 01:14:24,830
and they started developing models of reactions like test units like mass action models.

712
01:14:24,830 --> 01:14:29,960
And those same models are the ones that you know you can go to work departments theoretical, you know,

713
01:14:29,960 --> 01:14:42,120
emphatically that now that epidemiology or you can go to the old over here and they're working on the same types of equations,

714
01:14:42,120 --> 01:14:47,520
but they seem to miss something. So what is it that's missing? And how would you fix it?

715
01:14:47,520 --> 01:14:51,760
And it's through interactions with. How's it going? What have you done now?

716
01:14:51,760 --> 01:15:00,150
I haven't done it yet. That thinking, at least from my perspective, is as kind of developed.

717
01:15:00,150 --> 01:15:04,980
So what is the. I'll just go straight to the punch line because we're getting close to and the primary maths.

718
01:15:04,980 --> 01:15:09,240
You can do maths when things are homogenous and you can do maths when when I'm,

719
01:15:09,240 --> 01:15:13,290
well, mixed, you know, the standard test tube rate equation, things like that.

720
01:15:13,290 --> 01:15:18,030
You know, I take a test, you get I mix it. That's the same model that gives the all number on the BBC.

721
01:15:18,030 --> 01:15:27,420
When we hear about the R no record. It's the same model. It's all to do with when do I get something developed at the macro level or not?

722
01:15:27,420 --> 01:15:32,070
And it's all a test tube reaction between susceptibles infected, et cetera.

723
01:15:32,070 --> 01:15:37,320
In that case. But the same thing's true for models of the immune system, et cetera.

724
01:15:37,320 --> 01:15:40,840
So what is the solution so well or what is the.

725
01:15:40,840 --> 01:15:47,250
So actually, I'm going to show you the solution, and I'm also going to do this using two examples very quickly.

726
01:15:47,250 --> 01:15:53,550
It gives you the wrong idea about how things that match we've had amazing Claudia this morning.

727
01:15:53,550 --> 01:15:57,480
And you know how this organisation about, yeah, how does it develop?

728
01:15:57,480 --> 01:16:02,490
How did that amazing plume that Claudio showed? How did it just kind of was it nucleating?

729
01:16:02,490 --> 01:16:06,720
Did it just kind of peer? I mean, what is it and why didn't appear?

730
01:16:06,720 --> 01:16:10,140
If I'm looking at, why didn't it appear before? One thing is to watch it.

731
01:16:10,140 --> 01:16:14,640
Why didn't it appear 10 seconds before? You know what was wrong then?

732
01:16:14,640 --> 01:16:22,920
And why can't Max describe that? Or can it? So I took about that, and then I'm going to about just how long this stuff take to happen.

733
01:16:22,920 --> 01:16:28,200
I mean, you know, just just a standard thing. How long does stuff take to happen as a stand?

734
01:16:28,200 --> 01:16:31,950
How long does it take? How how long do you take to recover from COVID?

735
01:16:31,950 --> 01:16:37,680
How long does a reaction take? How long does any kind of complicated process take?

736
01:16:37,680 --> 01:16:42,850
We have maths for that, but it doesn't work. I'll show you what does it work?

737
01:16:42,850 --> 01:16:48,810
So these are the types of maths that you probably may have been faced with at some stage.

738
01:16:48,810 --> 01:16:55,470
You can have people over here come over here with kind of lattices and they'll show you things moving around like cellular automata.

739
01:16:55,470 --> 01:17:00,460
There are people now showing you like networks and, you know, networks developing all these kind of things.

740
01:17:00,460 --> 01:17:06,510
And and then they're these are back to a kind of standard reaction models.

741
01:17:06,510 --> 01:17:10,740
Mass action people show you those agent based models.

742
01:17:10,740 --> 01:17:19,710
And now, of course, the whole I, my group of physicists that we have a group of mathematician physicists and I people,

743
01:17:19,710 --> 01:17:23,790
you know, so they're always telling me, Oh, we're going to get 100 trillion parameters.

744
01:17:23,790 --> 01:17:27,630
Oh, that's great. That's fantastic. What insight can you possibly get?

745
01:17:27,630 --> 01:17:31,440
I can't even get insight from three. How can you get any insights?

746
01:17:31,440 --> 01:17:36,690
Because the answer is you can't. So there's a missing thing of maths that has a kind of simple model.

747
01:17:36,690 --> 01:17:40,500
Keep it simple. It's another one of those phrases. Keep it simple.

748
01:17:40,500 --> 01:17:46,260
Show simple, model simple, kind of twisting of the tubes that show DNA.

749
01:17:46,260 --> 01:17:53,730
Why can't we develop those kind of simple models? Comes down to laundry, so I want to show you.

750
01:17:53,730 --> 01:17:59,380
So this is is this correct along with this, but it's a simple it's.

751
01:17:59,380 --> 01:18:01,690
Think about this. It doesn't matter what you're looking at.

752
01:18:01,690 --> 01:18:08,630
You think of any assembly problem where something assembles by itself because you're doing it when you're going.

753
01:18:08,630 --> 01:18:16,990
I mean, imagine you've got some pile of laundry. This is just stuff available in some environment and it's it's going to self-assemble.

754
01:18:16,990 --> 01:18:23,020
Of course, if those items. So this is the one of the first things, this is the main thing that's wrong with nice models.

755
01:18:23,020 --> 01:18:27,490
Maths models assume the all that laundry is identical socks.

756
01:18:27,490 --> 01:18:31,870
I had a friend in college, you just bought identical socks and just do it because it doesn't matter.

757
01:18:31,870 --> 01:18:35,980
You never match them. They're just identical 100 pairs of identical socks.

758
01:18:35,980 --> 01:18:41,560
You just got a bag of 50. You don't have to assemble themselves already in pairs.

759
01:18:41,560 --> 01:18:44,900
You know, every two you find are a pair.

760
01:18:44,900 --> 01:18:51,470
But when you've got heterogeneous objects, you actually have the system in some sense, that's the kind of think as it assembles.

761
01:18:51,470 --> 01:18:57,350
And that takes time. I say it's an interesting question when you add heterogeneity to mass models of computer,

762
01:18:57,350 --> 01:19:02,510
which people don't do because it's hard to do the easy way to do it.

763
01:19:02,510 --> 01:19:09,890
But if you do that, then you know, I mean, any of us who came on a trip over here, you know, you go to events.

764
01:19:09,890 --> 01:19:13,880
I got to lady, you know, tonight, do I need a jacket with it in this?

765
01:19:13,880 --> 01:19:19,280
It's going to take one pair of shoes. I'll take those trousers. And what will that work them for the day?

766
01:19:19,280 --> 01:19:25,970
You know, you're thinking it's almost like the function then determines how things are assembling and it takes time.

767
01:19:25,970 --> 01:19:30,580
So. Again, I've only got a few minutes to tell you this,

768
01:19:30,580 --> 01:19:39,280
but we've worked out ways what we've been working on for the last 15, 20 years is ways to change the maths.

769
01:19:39,280 --> 01:19:52,320
To include heterogeneity, which seems to me to be the absolute number one problem with all modelling of biochemical systems.

770
01:19:52,320 --> 01:19:54,450
I'm not going to run these do all the maths,

771
01:19:54,450 --> 01:20:04,440
it's but the interesting thing is the time then that it takes and the size of the object and I don't mean like it could be an amyloids system.

772
01:20:04,440 --> 01:20:09,210
Yeah, OK. But that's getting into the details. What I also mean is, like more like a coherence.

773
01:20:09,210 --> 01:20:15,060
When is it the objects begin to cohere in some way they don't actually have to be next to each other.

774
01:20:15,060 --> 01:20:18,750
They just have to do something coherent in some coordinated way.

775
01:20:18,750 --> 01:20:24,830
And so we've been developing this maths of how you do that.

776
01:20:24,830 --> 01:20:33,490
How to describe that process. And the most interesting thing is you can get you get incredibly different growth curves.

777
01:20:33,490 --> 01:20:40,730
And the times at which things appear to change dramatically now, I've had no data,

778
01:20:40,730 --> 01:20:48,220
so I'm really hoping that at some stage by the time I know they'll be data from the biochemical area to look at these kind,

779
01:20:48,220 --> 01:20:55,300
of course, it's very difficult. So I had to go off and look at all kind of weird data like social media data, conflict data.

780
01:20:55,300 --> 01:21:00,130
And this is why, you know, I've never had a chance to interact with a lot of them.

781
01:21:00,130 --> 01:21:07,450
Talks with Dave and Lydia, but there is data in these other fields that act as proxies.

782
01:21:07,450 --> 01:21:13,720
Funnily enough, they act as proxies, particularly social media sites of how things organise form communities.

783
01:21:13,720 --> 01:21:15,850
Steve was talk about modular structure,

784
01:21:15,850 --> 01:21:23,140
some kind of like correlation from looking at the type of modular community structure the communities have loose connexions between them.

785
01:21:23,140 --> 01:21:25,990
It's exactly like a kind of what you might think.

786
01:21:25,990 --> 01:21:32,530
I might learn something from looking at a kind of a social media picture because there's real data there.

787
01:21:32,530 --> 01:21:38,950
And after all, we're all biological, some higher order level of organisation. But that might work.

788
01:21:38,950 --> 01:21:43,480
So I've got two questions here, first before lunch. What happens, for example? What's your intuition?

789
01:21:43,480 --> 01:21:47,470
If we're doing the laundry problem, we're assembling things and we start adding in more socks.

790
01:21:47,470 --> 01:21:53,740
So I got back to that laundry basket. I'm packing for the packing to come on the trip that someone's emptying the laundry, you know, is it?

791
01:21:53,740 --> 01:21:58,720
Does it just take longer? Is it do I get a richer set of clothes?

792
01:21:58,720 --> 01:22:04,660
Does it take the same? Is it shorter time because of the I've got lots of socks appearing that I could just stick in there?

793
01:22:04,660 --> 01:22:08,620
Well, the answer is it sort of depends. And that's the interesting bit. That's the interesting.

794
01:22:08,620 --> 01:22:13,090
This new maths, it's you can actually buy cheap by changing the flux, of course.

795
01:22:13,090 --> 01:22:17,350
None of none of you. Lot of us have environments that are closed.

796
01:22:17,350 --> 01:22:26,260
They all have fluctuating numbers of objects and heterogeneity, and it completely changes how the kind of coherence develops within the system.

797
01:22:26,260 --> 01:22:34,080
And I know I'm Typekit kind of in these generic terms. But I think that it's sufficiently generic that it's worth looking at.

798
01:22:34,080 --> 01:22:37,950
That's number one on my title before lunch, but so is this OK?

799
01:22:37,950 --> 01:22:44,890
That's how things develop. Things can develop, organise for good and for bad.

800
01:22:44,890 --> 01:22:51,130
What about the back? What about when you're trying to disassemble something, when you're trying to stop it from developing?

801
01:22:51,130 --> 01:22:56,680
Autoimmune disease, long COVID, whatever you want to do, whatever you want to talk about in some sense,

802
01:22:56,680 --> 01:23:00,970
is a red versus blue or blue versus red to flip it around.

803
01:23:00,970 --> 01:23:09,720
Red is trying to assemble something to maintain itself, and blue has got to work out some strategy or do something.

804
01:23:09,720 --> 01:23:17,790
Now, standard maths, because I'm meant to talk about what maths, what maths tells a standard maths tells us this.

805
01:23:17,790 --> 01:23:21,180
That the more, more blue you have, I mean, it's just, you know,

806
01:23:21,180 --> 01:23:26,010
it's like basic undergraduate chemistry, [INAUDIBLE] a lot of blue and a little bit of red.

807
01:23:26,010 --> 01:23:28,260
Well, that reaction's going to be over pretty quick.

808
01:23:28,260 --> 01:23:34,110
And if I've got even more blue, you know, I'm sure I get red and with some red, you know, it's going to be even quicker.

809
01:23:34,110 --> 01:23:38,220
So is the asymmetry between blue and red increases?

810
01:23:38,220 --> 01:23:45,990
You'd think that the reaction time would decrease worries me slightly because I know in my own mailbox, you know,

811
01:23:45,990 --> 01:23:52,990
when I've got a ton of unanswered mail, which I seem to have now, and I'm sure you all have, you know, it's like you can make a sweep through.

812
01:23:52,990 --> 01:23:58,350
But when you've got bits here and there and they're kind of locally clustered,

813
01:23:58,350 --> 01:24:04,540
but as they're spread out, part of me knows that actually that should probably take longer.

814
01:24:04,540 --> 01:24:11,290
And when we look at real data from long COVID, for example, this is so I wasn't in Glasgow, but Edinburgh, does that count?

815
01:24:11,290 --> 01:24:16,510
I'm sorry, it doesn't count. I know it's very separate and somebody told me that very separate cities.

816
01:24:16,510 --> 01:24:23,950
But Dave and I were at them with Nvidia and we were talking about this walking through the streets of Edinburgh just a few weeks ago.

817
01:24:23,950 --> 01:24:29,340
Why is it on Covid's log? I don't know.

818
01:24:29,340 --> 01:24:38,040
And the interesting thing is, when you look at the data of long Kobe being law, the models, the maths models sort of ignore it.

819
01:24:38,040 --> 01:24:43,920
Now this is on a bunch of ferrets. OK, so we'll give them, you know, we'll give them the benefit of the doubt.

820
01:24:43,920 --> 01:24:46,710
They didn't have more data and they didn't bend up their model.

821
01:24:46,710 --> 01:24:54,600
But there's existing data in the literature which exist, which suggests that you can't just fix things with the typical kind of reaction maths.

822
01:24:54,600 --> 01:25:01,890
This is typical reaction maths, even if you don't know about these. Every paper practically that you'll find we need to have this in it.

823
01:25:01,890 --> 01:25:05,280
Or like I saw the shortest, smallest model.

824
01:25:05,280 --> 01:25:08,970
Now for long, COVID has 15 equations, but they all look like this.

825
01:25:08,970 --> 01:25:13,710
And the only thing that matters is there's one derivative, which means it should be an exponential.

826
01:25:13,710 --> 01:25:18,450
And if it's an exponential, I should be able to plot to unlock linear plot and it should be a straight line.

827
01:25:18,450 --> 01:25:25,960
It ain't. It's longer it takes longer for things to happen, so I've got one slide.

828
01:25:25,960 --> 01:25:34,450
Left left ages ago when day was over with me in Miami and we were sitting out in somewhere,

829
01:25:34,450 --> 01:25:38,680
and I'm sure we're just having orange juice and we're having an orange, yes, mojito.

830
01:25:38,680 --> 01:25:42,850
And he asked at that time that was back in 2010.

831
01:25:42,850 --> 01:25:50,680
We go to Afghanistan, and that's the great thing. Insight into all these things, he was saying, Yeah, things wars take a long time.

832
01:25:50,680 --> 01:25:58,870
Don't lie. And it's like, Yeah, why do they actually do that? I don't know what set me off thinking about data on in terms of wars.

833
01:25:58,870 --> 01:26:01,990
It's the same thing as I just said to the asymmetry.

834
01:26:01,990 --> 01:26:09,370
When you think that a chemical reaction when when it's more and more asymmetric, blue versus red, it's going to be quicker and quicker.

835
01:26:09,370 --> 01:26:16,360
And yet when you plot out, it's a little bit off your topic that the plot out the asymmetry in conflicts and after all,

836
01:26:16,360 --> 01:26:25,150
the reaction is no more than the conflict in a red blue versus red could be the immune system versus cancer, the immune system versus cove.

837
01:26:25,150 --> 01:26:35,440
It could be any, any sort of competition. You plot them out and there's a high asymmetry in terms of the more blue them red.

838
01:26:35,440 --> 01:26:40,900
There's a moment when it's like the mailbox. I've got clusters, but I can't find them.

839
01:26:40,900 --> 01:26:47,500
So I'm so much blue, I've got so much blue that I can't find the red because I find myself every time I look,

840
01:26:47,500 --> 01:26:51,400
every time I'm a blue looking for stuff, I find more blue.

841
01:26:51,400 --> 01:26:57,570
I I'm fine red, red in these little clusters, and they kind of sit in pockets and they don't get touched.

842
01:26:57,570 --> 01:27:05,580
So we plotted this twenty nine, I remember the editor 2010, the editor published this 2010.

843
01:27:05,580 --> 01:27:15,880
The editor said, Oh yeah, those points are pretty out. Yes, 2010 was just ended, it hasn't even entered Afghanistan shoots up there.

844
01:27:15,880 --> 01:27:20,550
Colombia up the top that stopped.

845
01:27:20,550 --> 01:27:29,820
So it's not as crazy as it sounds that there's something to do with how living objects interact when they're asymmetric,

846
01:27:29,820 --> 01:27:35,970
that can make it last a really, really long time. So I lied, I got one more site.

847
01:27:35,970 --> 01:27:39,180
So we're now looking at this in terms of what's another thing. I've got the data.

848
01:27:39,180 --> 01:27:44,460
I'd love to get data from anybody, anybody who wants to, but I have to use all this other proxy data.

849
01:27:44,460 --> 01:27:49,740
So this is vaccine hesitancy. So this is something we're looking at now in terms of social media.

850
01:27:49,740 --> 01:27:54,920
Why is it when you got a ton of blue? They are CDC, everything.

851
01:27:54,920 --> 01:27:58,670
Each one of these dots is a community that could be like a few million people.

852
01:27:58,670 --> 01:28:02,840
Gates Foundation sitting up there, the red. So little anti-vaxxers.

853
01:28:02,840 --> 01:28:08,090
Anti-vax communities each read as an anti-vax community. Why is it the reaction takes so long?

854
01:28:08,090 --> 01:28:12,560
Why can't you get rid of them? It's not that you're trying to get rid of. There's a whole bunch of greens.

855
01:28:12,560 --> 01:28:18,320
It's like the background of the hearts and minds that kind of buried in the reaction.

856
01:28:18,320 --> 01:28:23,700
It's lasting a long time. It has an anomalous duration. So.

857
01:28:23,700 --> 01:28:28,710
Bring it to the end, because I know you would get so much. There's a problem last day was right.

858
01:28:28,710 --> 01:28:35,220
Twenty four years ago, there is a problem with that and you can't just carry on doing the same mess.

859
01:28:35,220 --> 01:28:38,310
You got to kind of grab it either grab the bull by the horns.

860
01:28:38,310 --> 01:28:44,430
And the hatching has started because that removes all the nice tricks and all the nice things.

861
01:28:44,430 --> 01:28:50,620
And yeah, OK. That's just the way it is got to include clustering community.

862
01:28:50,620 --> 01:28:55,390
So the same idea in terms of cohesion for any process that I would imagine that,

863
01:28:55,390 --> 01:29:01,400
you know about when I read textbooks, just stay beside me, read the textbooks I, I read them.

864
01:29:01,400 --> 01:29:05,950
It's not all these beautiful switches. It's like an electrical engineers diagram.

865
01:29:05,950 --> 01:29:13,240
And yet it cannot be like that. It's got to be collections of objects that may be a hundred, maybe five hundred and ninety six,

866
01:29:13,240 --> 01:29:17,950
but they switch and then seven thousand two hundred and twenty one don't.

867
01:29:17,950 --> 01:29:24,560
Why, you know, what's the kind of community? There must be some needs of scale, community level description.

868
01:29:24,560 --> 01:29:29,090
So anyway, I'm working on that, I'm so glad that it's retired and because we've got more time to talk about it.

869
01:29:29,090 --> 01:29:34,550
And the only way that I can think of the solution, what's the solution?

870
01:29:34,550 --> 01:29:41,480
Just get good friends who are really smart. So that's what Dave taught me.

871
01:29:41,480 --> 01:29:48,290
And so I'm really looking forward to this post phase where Dave and Lydia can meet up with us at any time,

872
01:29:48,290 --> 01:29:53,090
because these types of problems, I think, are the next frontier for modelling.

873
01:29:53,090 --> 01:29:58,760
I might see a trickle of people coming over from the other side of the road.

874
01:29:58,760 --> 01:30:05,530
Thanks very much. Thank you. That's.

875
01:30:05,530 --> 01:30:10,030
Donald, thanks very much, I really enjoyed that. So I wonder whether you could turn it upside down.

876
01:30:10,030 --> 01:30:15,790
Could it be that what's wrong with mass is actually what's wrong with biology in that we're coming from?

877
01:30:15,790 --> 01:30:25,990
We're expecting things to have one driver and one outcome from biology, maybe due to genetics, because you have a mutation, you have a phenotype.

878
01:30:25,990 --> 01:30:29,130
And then the realisation comes that.

879
01:30:29,130 --> 01:30:37,860
One gene, one protein, one consequence, you're either impotent or you're not important, and everything becomes necessary and sufficient,

880
01:30:37,860 --> 01:30:42,930
these two these diagrams which don't represent biology at all because you can have situations,

881
01:30:42,930 --> 01:30:47,100
for example, with the control element where any three things being turned on.

882
01:30:47,100 --> 01:30:52,860
Whichever combination they are gives you an outcome. It doesn't matter what the components are.

883
01:30:52,860 --> 01:30:57,510
And in a sense, what we're trying to model is the wrong thing. That's a fantastic point.

884
01:30:57,510 --> 01:31:09,200
Fantastic point. Yet it shifts the focus away from thermodynamics towards kinetic as well, I think, which absolutely should be on the kinetics.

885
01:31:09,200 --> 01:31:18,970
I'll ask a final question. So was this back to your first question, who should be doing something about it?

886
01:31:18,970 --> 01:31:25,610
Should I be waiting for the mother to be waiting a long time for that, actually?

887
01:31:25,610 --> 01:31:28,940
Or should we? Molecular biologist B,

888
01:31:28,940 --> 01:31:43,790
somehow attacking the mathematicians to get them to come was the only way I could get funding for this stuff is through Space Force,

889
01:31:43,790 --> 01:31:55,970
which, you know, I'm not going to sign any sizeable grant because they're interested in heterogeneous teams of machinery, humans.

890
01:31:55,970 --> 01:32:02,210
So you've got an element that's like a coin toss, maybe an element that has deterministic rules.

891
01:32:02,210 --> 01:32:09,200
Hopefully the machine and an element that doesn't take up any space but also has influence, which is the algorithm.

892
01:32:09,200 --> 01:32:14,300
So they're interested in teams. When is it that the team operates? Is it, you know, what is it?

893
01:32:14,300 --> 01:32:17,990
What's the magic mix? Is it? It's like cooking.

894
01:32:17,990 --> 01:32:20,660
It's like something like that. You know, what is the magic mix?

895
01:32:20,660 --> 01:32:26,870
And so they're really interested in that and how as you grow teams, it's an identical question as I take,

896
01:32:26,870 --> 01:32:32,510
you know, elements building blocks and put them together, and they're not identical.

897
01:32:32,510 --> 01:32:41,510
What how does the, you know, it's not changing from three to four, it's changing from fifty seven to six hundred and thirteen or something.

898
01:32:41,510 --> 01:32:45,890
It doesn't matter. It's 13, it could be 30. It's it's almost like another.

899
01:32:45,890 --> 01:32:52,820
It's like a meso level description in terms of groups.

900
01:32:52,820 --> 01:33:01,220
Yeah, I haven't answered the questions. So but I think you're going to be waiting a long time for the mathematicians to come across.

901
01:33:01,220 --> 01:33:11,144
Okay, thank you. Unfortunately, yeah, I'll say thanks to all of us because.