1 00:00:00,160 --> 00:00:00,400 Foreign. 2 00:00:05,760 --> 00:00:09,160 Welcome back to Translating Proteomics. Today I'm delighted to 3 00:00:09,160 --> 00:00:12,800 introduce you to Fiona Ginti, who is a senior principal scientist 4 00:00:12,800 --> 00:00:16,720 in precision diagnostics at the GE Healthcare Technology and 5 00:00:16,720 --> 00:00:20,680 Innovation Center. She is a leader in the development of spatial 6 00:00:20,680 --> 00:00:24,600 proteomics technologies and their application in precision diagnostics 7 00:00:24,600 --> 00:00:28,880 and medicine. And among the many other proof points of the importance 8 00:00:28,880 --> 00:00:33,000 of her work, Nature Methods named spatial proteomics the method 9 00:00:33,000 --> 00:00:37,200 of the year in 2024, highlighting the incredible potential 10 00:00:37,200 --> 00:00:41,120 for spatial proteomics technologies to expand our understanding 11 00:00:41,120 --> 00:00:45,200 of biological complexity. Additionally, Fiona is a leader on 12 00:00:45,200 --> 00:00:49,320 projects within NIH's Human Biomolecular Atlas Program or HUB 13 00:00:49,320 --> 00:00:53,080 MAP, which aims to create an open global atlas of the human body 14 00:00:53,320 --> 00:00:57,960 at scale cellular molecular resolution. Fiona, I'm really honored 15 00:00:57,960 --> 00:01:01,160 to have you here today and welcome to Translating Proteomics. 16 00:01:01,960 --> 00:01:05,640 Thank you, Parag. And it's great to see you again. We work a 17 00:01:05,640 --> 00:01:06,200 lot together. 18 00:01:06,600 --> 00:01:10,720 Yes. No, it's been such a fun, fun collaboration and I was thinking 19 00:01:10,720 --> 00:01:16,840 back to when we first met and I went back and I realized it was 20 00:01:16,840 --> 00:01:19,240 in February of 2013. 21 00:01:21,110 --> 00:01:21,430 Yes. 22 00:01:21,750 --> 00:01:26,470 So it was a while ago. And so you had come to visit us at Stanford 23 00:01:26,470 --> 00:01:31,070 at the Canary center and shared for the first time I'd ever 24 00:01:31,070 --> 00:01:37,910 seen this technology for studying multiplexed single cells. 25 00:01:38,390 --> 00:01:43,070 And I remember thinking, and I may have even said out loud, wow, 26 00:01:43,070 --> 00:01:49,100 that is the coolest thing I've ever seen. And then that led to an 27 00:01:49,100 --> 00:01:52,540 amazing partnership and multiple grants together and so much 28 00:01:52,540 --> 00:01:58,100 fun. I'd love to go back a little bit in time just to understand 29 00:01:58,100 --> 00:02:04,340 how you got to where you are and how you first got into biology 30 00:02:05,700 --> 00:02:11,620 and your early years studying bone biology at University College 31 00:02:11,620 --> 00:02:15,810 Cork. Maybe just share a little bit about your, the earliest 32 00:02:15,810 --> 00:02:16,490 part of your journey. 33 00:02:16,730 --> 00:02:20,050 Sure, yeah. Thank you for, yeah, going back down memory lane. 34 00:02:20,050 --> 00:02:26,090 It's always fun. So, yeah, I mean, you know, biology has always 35 00:02:26,090 --> 00:02:30,450 been of great interest. I think when, you know, at school level 36 00:02:30,450 --> 00:02:36,690 in high school, it's very much focused on, you know, zoology and 37 00:02:36,690 --> 00:02:41,180 bear form and, you know, you're not really thinking too far 38 00:02:41,180 --> 00:02:46,060 into the future, but it's, it's cool and it's interesting. And 39 00:02:46,780 --> 00:02:52,500 so I think, yeah, going to university College Galway, University 40 00:02:52,500 --> 00:02:57,820 of Galway we call now. I studied microbiology and just became, 41 00:02:58,220 --> 00:03:02,340 you know, passionate about bacteria and what bacteria could 42 00:03:02,340 --> 00:03:06,670 do. I just thought that was incredibly amazing that you could 43 00:03:06,670 --> 00:03:11,310 use bacteria for so many things and. But where I sort of got 44 00:03:11,310 --> 00:03:16,830 my interest more in the applied side of biology within. I 45 00:03:16,830 --> 00:03:21,270 did a course on nutrition for one year and just became fascinated 46 00:03:21,270 --> 00:03:26,670 by the idea of biomarkers. And at the time, I remember my professor, 47 00:03:26,670 --> 00:03:30,710 at the time, he, Professor Botchwell, he talked about nutrition 48 00:03:30,710 --> 00:03:33,670 biomarkers and I thought, what are they? That just sounds cool. 49 00:03:34,260 --> 00:03:40,340 So then I ended up going to Cork and working in bone, bone biomarkers. 50 00:03:40,980 --> 00:03:47,540 And that sort of led me to India and did a PhD in nutrition 51 00:03:48,100 --> 00:03:56,340 science and, you know, became more appreciative of what biomarkers 52 00:03:56,740 --> 00:04:04,110 can do in science, in biology and in clinical translation. So I 53 00:04:04,110 --> 00:04:07,990 just, Yeah, I think biomarkers has sort of been my north star and 54 00:04:08,870 --> 00:04:14,030 along the way. And then for. From there went and worked for Nestle 55 00:04:14,030 --> 00:04:19,030 for five years or two years and sold more of the applied side 56 00:04:19,030 --> 00:04:23,030 of, of nutrition, you know, how you food products and modify 57 00:04:23,030 --> 00:04:25,030 health. And you know. 58 00:04:25,030 --> 00:04:27,310 Well, that, that's already pretty interesting because I, I think 59 00:04:27,310 --> 00:04:30,970 most people, when they think about Nestle, they think candy bars. 60 00:04:31,930 --> 00:04:35,770 But of course Nestle has just an incredibly deep amount of research 61 00:04:35,850 --> 00:04:36,490 going on. 62 00:04:36,890 --> 00:04:40,050 Yeah, it was a. Yeah, it was an incredible research center. It 63 00:04:40,050 --> 00:04:44,330 is an incredible research center. And one thing I do remember 64 00:04:44,330 --> 00:04:48,810 when I went there for my job interview, there was chocolates there. 65 00:04:48,810 --> 00:04:51,290 You know, you could have like, you know, I don't know if they still 66 00:04:51,290 --> 00:04:54,170 do that, but that was a very impression, left a big impression 67 00:04:54,170 --> 00:04:54,650 on me. 68 00:04:54,890 --> 00:04:57,450 But when you're a young scientist, that is a huge draw. 69 00:04:58,650 --> 00:05:03,060 Yeah. So. But yeah, I think what was fascinating about the research 70 00:05:03,060 --> 00:05:08,500 center which stuck with me was it's multidisciplinary, it encompassed 71 00:05:08,820 --> 00:05:11,980 nutrition, whether it was food science, food technology, as well 72 00:05:11,980 --> 00:05:15,540 as the whole sort of field of biomarkers and how you could measure 73 00:05:15,540 --> 00:05:18,700 the effects on health. So I think that is something that stuck 74 00:05:18,700 --> 00:05:23,620 with me too, is just the team science aspects of how important 75 00:05:23,620 --> 00:05:25,540 it is to work with other disciplines. 76 00:05:25,900 --> 00:05:30,180 I'm so curious about how you went from first studying and learning 77 00:05:30,180 --> 00:05:35,940 about fundamental aspects of how we detect what's happening in 78 00:05:35,940 --> 00:05:40,780 the body and how we alter that with nutritional programs to then 79 00:05:41,819 --> 00:05:46,540 building entirely new technologies. And so I'm curious 80 00:05:46,540 --> 00:05:51,740 about, was that an explicit mind shift or that you like, wow, 81 00:05:51,740 --> 00:05:56,110 I really wish we had better tools or how did that transition 82 00:05:56,110 --> 00:05:56,670 come about? 83 00:05:57,790 --> 00:06:03,790 Yeah, so I think it was, it was gradual. So I. One thing I came 84 00:06:03,790 --> 00:06:08,669 to appreciate through nutrition research and still do is 85 00:06:08,669 --> 00:06:13,910 how, you know, it's not one measurement that determines your 86 00:06:13,910 --> 00:06:17,750 health, it's not one nutrient that determines your health. You 87 00:06:17,750 --> 00:06:22,250 know, sort of the balance of nutrition, different, you know, different 88 00:06:22,650 --> 00:06:26,610 vitamins, you know, so forth fat, protein. So it's a sort of a 89 00:06:26,610 --> 00:06:30,410 system of systems. And then the Measurement tools are so important. 90 00:06:30,730 --> 00:06:34,450 So sort of like my mindset was already kind of like, okay, all these 91 00:06:34,450 --> 00:06:38,970 things have to work together to produce a perfect outcome or near 92 00:06:38,970 --> 00:06:42,690 whatever. Near a perfect outcome. But I think the thing that's 93 00:06:42,690 --> 00:06:46,780 hard when as a biologist you're, you know, for me this is 94 00:06:46,780 --> 00:06:50,020 what kind of drove me, I think into the new technology space was 95 00:06:50,020 --> 00:06:53,740 that you are, you know, you're, you're, you're, you're limited 96 00:06:53,740 --> 00:06:59,460 to the tools that you can buy off the shelf. And so, and you always 97 00:06:59,460 --> 00:07:02,820 just have more questions, you know, how, you know, if I could, 98 00:07:02,820 --> 00:07:05,900 one more measurement, I could do this. But you're still limited 99 00:07:05,900 --> 00:07:10,180 by, you know, what's available. But an opportunity came 100 00:07:10,180 --> 00:07:15,700 to move to the US and moved to ge. And yeah, at the time it was 101 00:07:15,700 --> 00:07:19,460 like when I first started considering ge, I didn't realize 102 00:07:19,460 --> 00:07:19,780 what. 103 00:07:20,180 --> 00:07:22,340 Right. Again, when people think about ge, they think light 104 00:07:22,340 --> 00:07:26,780 bulbs, but they forget that GE has this whole other aspect to it 105 00:07:26,780 --> 00:07:33,420 with MRI scanners and so much wider. And also where you went in 106 00:07:33,420 --> 00:07:38,980 GEG research. Most people are probably not familiar with GE research. 107 00:07:39,300 --> 00:07:44,090 So tell us a little bit about what GE research was and its mission 108 00:07:44,090 --> 00:07:49,690 and how it's sort of, sort of academic, but also sort of industry 109 00:07:49,850 --> 00:07:51,050 and living in the middle. 110 00:07:51,450 --> 00:07:54,810 Yeah, I think it's a very unique model. I mean often it's really 111 00:07:54,810 --> 00:07:58,810 compared to Bell Labs, you know, was one of the last industrial 112 00:07:58,810 --> 00:08:04,010 labs in the country. You know, it's thousand scientists, engineers 113 00:08:04,010 --> 00:08:08,610 on site joined with the, I joined the biology biosciences group 114 00:08:08,610 --> 00:08:14,370 which GE had just bought Amersham and you know, known for 115 00:08:14,610 --> 00:08:17,890 pharma tools, molecular imaging. And so we were building 116 00:08:19,090 --> 00:08:23,890 a new department and but the fact that there were so many skills 117 00:08:23,890 --> 00:08:26,930 that came from other backgrounds that were non biology 118 00:08:27,250 --> 00:08:32,130 actually started to bring together a lot of new ideas. We were 119 00:08:33,720 --> 00:08:38,360 tasked with, our mission was to, to do for pathology what had 120 00:08:38,360 --> 00:08:43,080 been done for radiology in terms of going from film to digital 121 00:08:43,560 --> 00:08:48,000 radiology. And so our, our mission was to go from slides to 122 00:08:48,000 --> 00:08:52,440 digital pathology. Because the problem with pathology as you know, 123 00:08:52,440 --> 00:08:56,120 is, you know, you've got one piece of tissue on a glass slide 124 00:08:56,120 --> 00:09:00,120 and that was used by pathologists to make the diagnosis. 125 00:09:00,430 --> 00:09:03,310 And then if there needed to be a second diagnosis or referral, then 126 00:09:03,310 --> 00:09:06,670 that glass slide had to go be undelivered. Essentially. 127 00:09:07,150 --> 00:09:09,550 People were physically mailing around slides. 128 00:09:09,550 --> 00:09:12,510 Exactly. You know, when you think about it, I mean, I know it 129 00:09:12,510 --> 00:09:15,550 still happens and you know, there's different. But at the time 130 00:09:15,550 --> 00:09:20,510 that was the only, pretty much the only way it was done. And, yeah, 131 00:09:20,510 --> 00:09:24,910 I mean, and there's so much data in, in, in an image. You know, 132 00:09:24,910 --> 00:09:28,380 that was the thing, I think, that was compelling for us. On the 133 00:09:28,380 --> 00:09:30,900 one hand, we sort of take the engineering view. Okay, we're going 134 00:09:30,900 --> 00:09:33,220 to scan this slide and we're going to do it really quickly. I 135 00:09:33,220 --> 00:09:36,500 mean, you know, we knew what we needed to do for it to be successful. 136 00:09:36,500 --> 00:09:40,700 It had to be scanned in a very short amount of time. And then all 137 00:09:40,700 --> 00:09:42,740 the QC around that and, you know, there was a big. 138 00:09:42,740 --> 00:09:46,420 Team who was, who was on that team, were there because I'm, I'm. 139 00:09:46,660 --> 00:09:52,420 And what. So timeline wise, this is 2009. 2000 or earlier? 140 00:09:53,080 --> 00:09:57,080 It's 2005, actually. Yeah. So. 141 00:09:57,560 --> 00:10:01,840 Wow. And so, so just even the, the technology to take a picture 142 00:10:01,840 --> 00:10:05,720 and digitize it was not what we're accustomed to today. 143 00:10:06,360 --> 00:10:11,240 No. And, you know, the team was just like, I mean, it blew my 144 00:10:11,240 --> 00:10:15,320 mind, you know, that the. None of us were pathologists. We were 145 00:10:15,320 --> 00:10:18,760 not trained in pathology. There were certainly people on the 146 00:10:18,760 --> 00:10:22,970 team who had expertise and so they were able to, you know, sort 147 00:10:22,970 --> 00:10:27,090 of help guide at least the initial questions. But, you know, 148 00:10:27,090 --> 00:10:32,730 we worked with pathologists who helped guide us in terms of what 149 00:10:32,730 --> 00:10:37,090 they needed. And we watched their workflows and went on site 150 00:10:37,090 --> 00:10:41,010 to look how they, how they did things, what the, you know, all the 151 00:10:41,010 --> 00:10:44,810 barriers along the whole workflow, how long it took to read 152 00:10:44,810 --> 00:10:48,170 a slide. And they're just very simple things that, you know, just 153 00:10:48,170 --> 00:10:52,630 get. Gave the team the ideas that needed to be implemented. So, 154 00:10:53,430 --> 00:10:59,590 you know, and I think that sort of quickly evolved from just 155 00:11:00,070 --> 00:11:05,590 digitizing the slide to the imaging of the slide of the proteins 156 00:11:05,590 --> 00:11:09,110 and the information that were sort of locked within the image. 157 00:11:09,110 --> 00:11:12,470 And that sort of was the eureka for me at first, you know, 158 00:11:12,470 --> 00:11:15,990 I was just fascinated. Okay, this is cool. I had never worked 159 00:11:15,990 --> 00:11:19,490 in pathology, but the fact that this was, you know, these are, 160 00:11:20,130 --> 00:11:23,970 you know, they're unvisual persons, so you can see there's information 161 00:11:24,130 --> 00:11:25,810 there, but, you know, what does it mean? 162 00:11:26,290 --> 00:11:29,610 Can you maybe walk me through a little bit that, that transition 163 00:11:29,610 --> 00:11:36,210 from I'm just trying to, you know, make pathology digital to I, 164 00:11:36,290 --> 00:11:40,690 I actually see that there's this space of spatial biology and 165 00:11:41,010 --> 00:11:44,090 molecular spatial biology that happened in your. 166 00:11:44,090 --> 00:11:48,240 Head once we figured out the multiplexing part, which is the issue 167 00:11:48,240 --> 00:11:52,000 sustaining on the same sample. I mean, a lot of that work is going 168 00:11:52,000 --> 00:11:57,800 back to 2008. And, you know, we were certainly, you know, just 169 00:11:57,800 --> 00:12:01,440 didn't know what to expect in terms of how it would look. The early 170 00:12:01,440 --> 00:12:05,440 images were, you Know, not great quality. And so once we kind 171 00:12:05,440 --> 00:12:05,760 of got. 172 00:12:06,000 --> 00:12:11,520 So it was already 2008 that you were using protein fluorescence 173 00:12:12,550 --> 00:12:14,950 fluorescent imaging on your pathology data. 174 00:12:15,270 --> 00:12:18,670 Yeah, I think we might have even produced the first images in 175 00:12:18,670 --> 00:12:23,350 2007. I mean, we were doing several things at the same time and 176 00:12:23,830 --> 00:12:26,670 you know, some of the team were focused on the chemistry, you 177 00:12:26,670 --> 00:12:32,470 know, how do we exploit the fact that these dyes are not stable. 178 00:12:32,710 --> 00:12:41,350 And so we had an amazing team. Anuxud, you know, he came up with 179 00:12:41,350 --> 00:12:44,870 the original concept of how to, you know, inactivate the dyes. 180 00:12:45,030 --> 00:12:48,910 And so we, you know, we tried other methods too. We, you know, 181 00:12:48,910 --> 00:12:52,870 we tried light, we tried, you know, different compositions. But 182 00:12:52,870 --> 00:12:57,710 the, the dye, the chemical inactivation was the most conducive 183 00:12:57,710 --> 00:13:01,350 to doing a whole slide, you know, versus light, you know, it's 184 00:13:01,350 --> 00:13:05,720 more focal. And so, you know, those first experiments, you know, 185 00:13:05,720 --> 00:13:09,760 we, we had to prove out how many times could we iterate over, 186 00:13:09,840 --> 00:13:14,920 you know, 100 rounds of staining. And I remember another 187 00:13:14,920 --> 00:13:20,720 team member, Max Seal, he, he, he said, let's just keep going. I'm 188 00:13:20,720 --> 00:13:22,960 just going to keep staining and staining and staining. And we 189 00:13:22,960 --> 00:13:27,400 were like, okay, are you done yet? Are you done yet? And so, and 190 00:13:27,400 --> 00:13:31,610 then I'm another team member, Chris Savinsky, me wanted again, 191 00:13:31,610 --> 00:13:34,970 he had, you know, the early vision. Hey, let's get immune cells 192 00:13:34,970 --> 00:13:38,850 in there as well. This was, you know, maybe 2009 or 10. 193 00:13:39,410 --> 00:13:42,890 Oh my gosh. So I had no idea that this was, this was happening 194 00:13:42,890 --> 00:13:49,570 that early because I feel like your first multi omics paper was 195 00:13:49,570 --> 00:13:53,930 the 2013 Gerdus paper, is that right? 196 00:13:53,930 --> 00:13:58,980 That's right. So we actually. Yeah, so Michael Gurde, he's another 197 00:13:58,980 --> 00:14:04,420 critical team member, he had the vision around combining it with 198 00:14:04,420 --> 00:14:10,380 fish and amongst other things. So we had a lot of really great brainstorming 199 00:14:10,380 --> 00:14:14,220 during that time and then we were trying to couple it with the 200 00:14:14,220 --> 00:14:19,260 imaging workflow. So there was always this what if, but we need 201 00:14:19,260 --> 00:14:22,860 to do this, but we need to register the images. And so we had, 202 00:14:23,230 --> 00:14:27,150 you know, another engineer in the team, Kevin Kenney, who was helping 203 00:14:27,150 --> 00:14:31,070 us register all the images. And then there was the auto fluorescence. 204 00:14:31,070 --> 00:14:35,030 So we knew that that was going to be a co founder. So yeah, so the 205 00:14:35,030 --> 00:14:41,030 first data set I think probably was about 2008, 2009 that 206 00:14:41,030 --> 00:14:42,990 we could actually do some analysis. 207 00:14:42,990 --> 00:14:48,430 On just to just explain for a second, sort of come forward in time 208 00:14:49,260 --> 00:14:53,980 just what the general workflow and approach of cell dive multiomics 209 00:14:54,380 --> 00:14:58,900 is and then we'll go back and dive Into a little bit more how we 210 00:14:58,900 --> 00:15:02,620 got there, Because I think this concept of iterative staining 211 00:15:02,620 --> 00:15:08,820 and dye inactivation, how does cell dive actually make a multiplex 212 00:15:08,820 --> 00:15:09,900 single cell measurement? 213 00:15:11,340 --> 00:15:16,610 So in conventional immunohistochemistry, staining standard, 214 00:15:16,610 --> 00:15:20,930 what you do is use an antibody that's labeled, has got a tag on 215 00:15:20,930 --> 00:15:28,050 it, some kind of signal that can be detected by light. And the 216 00:15:28,050 --> 00:15:31,650 basic principle is you use an antibody, but attached to a fluorescent 217 00:15:31,650 --> 00:15:36,450 dye, and then that antibody binds to the protein on the tissue, 218 00:15:36,530 --> 00:15:41,330 so it lights up wherever that protein is on the tissue, somewhere 219 00:15:41,570 --> 00:15:47,740 in the cell or multiple cells. And so that gives you one photograph 220 00:15:47,740 --> 00:15:53,340 or one picture. But then you can get rid of the fluorescence dye 221 00:15:53,420 --> 00:15:57,420 that, you know, the tag that's lighting up that cell. And then you 222 00:15:57,420 --> 00:16:03,260 can add another antibody with maybe the same dye or a different 223 00:16:03,260 --> 00:16:06,740 dye. So you're basically taking multiple pictures of the same 224 00:16:06,740 --> 00:16:11,500 sample in an interest of workflow, and you're inactivating 225 00:16:11,500 --> 00:16:16,020 the dye every time it dies. And that way you're building up a 226 00:16:16,020 --> 00:16:19,260 montage of pictures of the same sample. 227 00:16:20,540 --> 00:16:24,740 And then because you register them, you're then able to say, I 228 00:16:24,740 --> 00:16:28,620 had this marker, There was this much of it in this particular 229 00:16:29,260 --> 00:16:33,060 cell at this coordinate, and then marker 2 and marker 3 and marker 230 00:16:33,060 --> 00:16:36,800 4. So you build up this data cube where you've got, you know, 231 00:16:37,520 --> 00:16:43,040 some large number, 40, 50, 200 different proteins measured on a 232 00:16:43,040 --> 00:16:44,240 per cell basis. 233 00:16:44,880 --> 00:16:48,520 Exactly. And I think that was sort of one of our key things that 234 00:16:48,520 --> 00:16:53,000 we got into early, because there have been other multiplexing 235 00:16:53,000 --> 00:16:56,280 methods before, is that did things a little differently, either 236 00:16:56,280 --> 00:17:01,520 with light or had a chemical removal step. But we, one, we needed 237 00:17:01,520 --> 00:17:04,950 to know we needed a dental process because we needed. Sample 238 00:17:04,950 --> 00:17:09,270 needed to survive, you know, 30, 40 times where it was going to 239 00:17:09,270 --> 00:17:12,990 be stained and restained. So it had to be a gentle process. And 240 00:17:12,990 --> 00:17:16,950 then secondly, we didn't want to. We wanted to work at a single 241 00:17:16,950 --> 00:17:22,350 cell level rather than just cover kind of a lot of data, pixel 242 00:17:22,350 --> 00:17:24,150 level. You know, that would be hard. 243 00:17:24,470 --> 00:17:28,990 That was an explicit choice to say, we actually want to view this 244 00:17:28,990 --> 00:17:31,990 from a cell perspective, not just a pixel perspective. 245 00:17:32,490 --> 00:17:36,450 Exactly. And I think the first thing that drove us as well was we 246 00:17:36,450 --> 00:17:40,730 were really thinking from a clinical point of view at the very 247 00:17:40,730 --> 00:17:46,090 beginning. Meaning, you know, if you take breast cancer, for example, 248 00:17:46,170 --> 00:17:50,370 you know, that takes, you know, her to estrogen receptor. There's 249 00:17:50,370 --> 00:17:55,850 three or four proteins that are important from a diagnostic perspective. 250 00:17:55,850 --> 00:17:59,170 So our view, at least in the beginning, was, well, what about, 251 00:17:59,170 --> 00:18:01,850 wouldn't it be cool if we could do all that on the same slide. 252 00:18:01,850 --> 00:18:05,630 That's not actually where we ended up, but that was our mindset 253 00:18:05,630 --> 00:18:09,470 that we wanted to measure at a cell level. It wasn't even at the 254 00:18:09,470 --> 00:18:12,630 very beginning, a single cell level. That came a little later. 255 00:18:12,950 --> 00:18:17,190 We were just looking at sort of overall signal within a sample. 256 00:18:17,270 --> 00:18:21,390 You know, was it above a threshold or below a threshold? And 257 00:18:21,390 --> 00:18:24,870 then. But then we realized, oh, no, we need to actually go single 258 00:18:24,870 --> 00:18:25,350 cell. 259 00:18:25,750 --> 00:18:28,230 So that's really interesting because there was a. There's a moment 260 00:18:28,230 --> 00:18:32,150 in time, and I think my lab was. Was converging around this moment 261 00:18:32,150 --> 00:18:37,790 in time where it felt like as a community, we came to appreciate 262 00:18:38,670 --> 00:18:44,710 tumor heterogeneity. And so I'm wondering if this. If it was 263 00:18:44,710 --> 00:18:48,750 that moment where you're like, oh, actually, it's not just the total 264 00:18:48,750 --> 00:18:53,430 signal. We've got all these different variations from cell to 265 00:18:53,430 --> 00:18:54,830 cell that we want to capture. 266 00:18:55,870 --> 00:18:59,870 Yeah, I think that was. We once. We sort of. I think that was 267 00:18:59,950 --> 00:19:04,870 sort of an important learning point for us once we stopped only 268 00:19:05,030 --> 00:19:09,750 thinking about trying to replicate, like, a current workflow, 269 00:19:09,990 --> 00:19:15,350 which at the time was the pathologist's workflow of whether 270 00:19:15,430 --> 00:19:19,190 a sample was positive or negative for a given biomarker. And 271 00:19:19,349 --> 00:19:23,590 we started to work with. Collaboratively with Eli lilly in 272 00:19:23,830 --> 00:19:29,620 about 2010. That was another awesome leap for us because so now 273 00:19:29,620 --> 00:19:33,420 we were okay. We had sort of rudimentary tools developed, but 274 00:19:33,580 --> 00:19:37,460 we. We needed to show them off. We needed people to sort of 275 00:19:37,460 --> 00:19:41,740 see, try to understand from a pharma perspective what was important 276 00:19:42,380 --> 00:19:46,700 from, you know, drug response perspective. And heterogeneity sort 277 00:19:46,700 --> 00:19:52,780 of slowly evolved out of that. But once we knew how to do single 278 00:19:52,780 --> 00:19:56,990 cell analysis, we were able to start asking those questions because 279 00:19:56,990 --> 00:20:00,150 you could see it in the images. It's not like, you know, 280 00:20:00,150 --> 00:20:03,190 sort of start to look at these stains or relate with each other. 281 00:20:03,190 --> 00:20:06,510 You're like, oh, my. There's a cluster of cells there. And there's. 282 00:20:06,510 --> 00:20:11,070 There's sort of similar cluster cells there. And, oh, there's 283 00:20:11,070 --> 00:20:15,950 immune cells there too. And so you start to think about things in 284 00:20:15,950 --> 00:20:19,350 a little different way that we'd ever kind of thought of before 285 00:20:19,590 --> 00:20:22,830 this. And. And then how do you quantify this? It's like that was 286 00:20:22,830 --> 00:20:29,030 our next kind of conundrum. How do you convert what you see into 287 00:20:29,030 --> 00:20:29,590 a number? 288 00:20:30,630 --> 00:20:33,750 That's right. Because the difference between, hey, percent 289 00:20:33,830 --> 00:20:37,350 positive, there's no quantitation. There's just like, 290 00:20:37,830 --> 00:20:42,790 is it above threshold or below threshold or percent positivity. 291 00:20:43,350 --> 00:20:46,310 It's a really straightforward quantification. 292 00:20:47,160 --> 00:20:51,800 Yeah, exactly. And I think that's where we were still kind of 293 00:20:51,960 --> 00:20:55,560 thinking, I suppose, bound by thresholds, you know, that being 294 00:20:55,560 --> 00:21:00,120 the only way to understand like the biology, you know, above 295 00:21:00,120 --> 00:21:03,480 this, then it means this. If it's below this, it means that. And 296 00:21:04,120 --> 00:21:05,800 so we're able to segment every. 297 00:21:05,800 --> 00:21:10,720 Cell in the image, which is a non trivial problem. Like that is 298 00:21:10,720 --> 00:21:13,400 an incredibly hard problem. And for some technologies today, 299 00:21:14,030 --> 00:21:16,030 that is still a mostly unsolved problem. 300 00:21:16,750 --> 00:21:21,230 Yeah, that was really challenging. We tried to do something 301 00:21:22,030 --> 00:21:27,030 fairly straightforward, which was to use a membrane marker. So 302 00:21:27,030 --> 00:21:33,470 we had DAPI for the nucleus, we had cytokeratin and we had the 303 00:21:33,950 --> 00:21:38,710 membrane marker. And so we had sort of the key elements of a cell, 304 00:21:38,710 --> 00:21:42,460 if you like. I mean it wasn't perfect, but we were able to at least 305 00:21:42,460 --> 00:21:48,980 segment the cells. And what we realized in doing so was that you 306 00:21:48,980 --> 00:21:56,180 could conserve the spatial coordinate and generate an ID for 307 00:21:56,180 --> 00:22:00,740 each cell. But we still didn't know what to do next. We had a scientist 308 00:22:00,740 --> 00:22:04,540 on the team who was working with another scientist. This is often 309 00:22:04,540 --> 00:22:09,270 how things work in that environment. Who had worked in geospatial 310 00:22:10,630 --> 00:22:15,190 map. Colin McCall, I was his name. He, he, he said, okay, let's 311 00:22:15,190 --> 00:22:18,470 talk to Tom, because Tom has done this geospatial analysis, I 312 00:22:18,470 --> 00:22:21,510 think that might help us. So we explained to Tom what we wanted 313 00:22:21,510 --> 00:22:25,270 to do, that we wanted to map the cells and look at their, you 314 00:22:25,270 --> 00:22:28,750 know, the neighborhoods or see what they were co localized with. 315 00:22:28,750 --> 00:22:32,030 And Tom said, oh, that's easy. You know, that's, we do that all 316 00:22:32,030 --> 00:22:32,470 the time. 317 00:22:33,480 --> 00:22:37,240 We frequently register thousands of images together. No 318 00:22:37,240 --> 00:22:37,720 big deal. 319 00:22:38,040 --> 00:22:41,480 Yeah. And so we were like, okay, here's our data set. And then 320 00:22:41,960 --> 00:22:45,320 he came back like the next day and showed us, you know, again, very 321 00:22:45,320 --> 00:22:50,960 rudimentary co localization of the different cell types. And we 322 00:22:50,960 --> 00:22:55,880 were like, oh wow, that's, that's what we wanted. And so that 323 00:22:55,880 --> 00:22:59,760 sort of was our, that was our step one. But they were all really 324 00:22:59,760 --> 00:23:01,220 long steps. Yes. 325 00:23:01,380 --> 00:23:04,820 And was there, was there a moment there and, and when was it 326 00:23:05,060 --> 00:23:09,540 that you sort of got that data set that you're like, wow, this, 327 00:23:09,940 --> 00:23:10,900 this works? 328 00:23:11,540 --> 00:23:15,900 I think one of the Eureka moments was we were working with 329 00:23:15,900 --> 00:23:20,580 Eli Lilly at the time. They had given us samples that, you know, 330 00:23:20,580 --> 00:23:25,940 were drug treated in some way. So you know, in theory there was 331 00:23:26,540 --> 00:23:30,100 going to be a heterogeneous response. It was, you know, we kind 332 00:23:30,100 --> 00:23:33,460 of knew that going in. And the question was whether we could quantify 333 00:23:33,460 --> 00:23:36,940 that or not. Taking all the learnings that we had around single 334 00:23:36,940 --> 00:23:41,180 cell segmentation, mapping them Back to the image. And then 335 00:23:41,980 --> 00:23:45,820 I remember another one of our team members, Megan Rothman, she 336 00:23:46,540 --> 00:23:51,740 had applied a clustering K means clustering method to identify 337 00:23:51,900 --> 00:23:55,730 locations. So to find the common groupings of cells or cells 338 00:23:55,730 --> 00:23:59,290 of common characteristics. And then because we have the spatial 339 00:23:59,290 --> 00:24:02,610 location, we could see where they map back to. On the. On the 340 00:24:02,610 --> 00:24:05,970 sample. And sort of was amazing. Like, there was sort of. 341 00:24:05,970 --> 00:24:09,650 You could see this gradation of, you know, where there was, you 342 00:24:09,650 --> 00:24:17,050 know, pockets of, you know, either response or cell types. And 343 00:24:18,170 --> 00:24:22,350 so we thought, wow, that. That looks. That looks really cool. And 344 00:24:22,350 --> 00:24:26,070 then I remember we showed this to Lily and they were, wow, this 345 00:24:26,070 --> 00:24:30,070 is the Holy Grail. We were like, oh, if it's the Holy Grail, 346 00:24:31,110 --> 00:24:35,030 that's amazing. If that's the Holy Grail. And I remember Chris 347 00:24:35,030 --> 00:24:37,590 Vinsky coming in one day. He says, hey, we really need to start 348 00:24:37,590 --> 00:24:40,790 adding immune cells into our workflow. 349 00:24:40,870 --> 00:24:44,950 Because you were focused so much on the epithelial cells, probably 350 00:24:44,950 --> 00:24:48,130 not even the endothelial cells. You were focused on the cancer 351 00:24:48,130 --> 00:24:48,570 cells. 352 00:24:48,890 --> 00:24:52,650 Exactly. And, you know, okay, let's put them in as well. And then 353 00:24:52,650 --> 00:24:56,690 we started getting these beautiful images. You know, Chris 354 00:24:56,690 --> 00:24:59,370 was doing this colon cancer study which ended up being part of 355 00:24:59,370 --> 00:25:03,970 the PNAS paper that published in 2013. And we could see, oh, this 356 00:25:03,970 --> 00:25:06,730 is incredible. And an immunoscore had just come out at 357 00:25:06,730 --> 00:25:10,770 the time as well from Jerome Galan. And so we're like, oh, this 358 00:25:10,770 --> 00:25:14,650 is actually really big. This, you know, we need to understand. 359 00:25:14,730 --> 00:25:17,690 We need to have immune themselves as part of every study 360 00:25:17,690 --> 00:25:22,410 that we do going forward. And so, yeah, we were. We were. We were 361 00:25:22,410 --> 00:25:25,450 having fun with this, but we also knew we were. We were working 362 00:25:25,450 --> 00:25:30,290 towards. We needed to develop something, you know, for GE as well. 363 00:25:30,290 --> 00:25:35,370 You know, we. Building a workflow, and it needed to. 364 00:25:35,370 --> 00:25:37,930 Evolve from being a, hey, look, we can do it into something 365 00:25:37,930 --> 00:25:41,410 that could be more mature and productized and. 366 00:25:42,570 --> 00:25:45,650 Exactly. And I think that that was like another transition point 367 00:25:45,650 --> 00:25:52,570 for us, where we started then to work with Clarion, which GE had 368 00:25:52,650 --> 00:26:02,210 bought in 2011 or 2012. And that was sort of our technology transfer 369 00:26:02,210 --> 00:26:05,770 moment that we. We started to sort of think more practically in 370 00:26:05,770 --> 00:26:12,220 terms of how this could work in the clinical context. And that 371 00:26:12,220 --> 00:26:17,300 was also very important from a QC perspective, because now we really 372 00:26:17,300 --> 00:26:22,900 have to start thinking about workflow. And the team members, Alex 373 00:26:22,900 --> 00:26:28,620 and Bob, and they really guided the team in terms of how do 374 00:26:28,620 --> 00:26:33,220 we make something that can be used in a very robust, reproducible 375 00:26:33,220 --> 00:26:35,300 context. So that was another big transition. 376 00:26:35,780 --> 00:26:38,580 It sounds like there was a first vision which was just how do 377 00:26:38,580 --> 00:26:43,080 we make pathology digital? And then that evolved into how do we 378 00:26:43,080 --> 00:26:48,800 get more information density into, into pathology? And then you 379 00:26:49,600 --> 00:26:53,360 accomplish that and then had sort of two parallel paths. One was 380 00:26:53,680 --> 00:26:57,480 let's use this and learn things and do science. And then the 381 00:26:57,480 --> 00:27:00,240 other was let's continue to advance and mature the technology 382 00:27:01,040 --> 00:27:01,680 as well. 383 00:27:02,160 --> 00:27:05,040 Yeah, exactly. I think. And that was, you know, that was sort 384 00:27:05,040 --> 00:27:10,050 of, you know, moments where we were able to do both or. And then, 385 00:27:10,050 --> 00:27:14,490 you know, we started to work with Michael Bernie started working 386 00:27:14,490 --> 00:27:18,370 Vanderbilt and we got one of our first NIH grants with, with the 387 00:27:18,370 --> 00:27:23,290 team there and Bob Coffey. And, you know, it was a matter then 388 00:27:23,290 --> 00:27:26,970 of being able to prove out the value for the research community 389 00:27:26,970 --> 00:27:31,250 as well. That was one of the first single cell analysis programs 390 00:27:31,250 --> 00:27:35,770 that the NIH was running. I think that was maybe 2014, 2015, 391 00:27:36,340 --> 00:27:41,620 but the 2013 paper, you know, that was our, you know, our watershed 392 00:27:41,620 --> 00:27:44,900 moment in a sense of getting the paper out there. But that took, 393 00:27:45,060 --> 00:27:48,980 you know, a good three years to bring all those pieces together. 394 00:27:49,460 --> 00:27:54,660 Wow. So from beginning to end, it sounds like it was eight years 395 00:27:55,620 --> 00:28:02,260 from the start of the journey, 2005, 2006 to that watershed moment. 396 00:28:03,250 --> 00:28:04,130 Yeah, exactly. 397 00:28:04,210 --> 00:28:08,770 If you think just in the last sort of three years, what, what studies, 398 00:28:08,770 --> 00:28:11,250 what projects have you been most excited about? 399 00:28:12,530 --> 00:28:15,530 I think that one of the biggest, most exciting areas that 400 00:28:15,530 --> 00:28:19,490 we've worked on in the last, actually it's been seven years or 401 00:28:19,490 --> 00:28:23,090 six to seven years, is the HUBMAP program, which is the human 402 00:28:23,090 --> 00:28:29,110 Biomolecular Mapping project. This is a massive effort that sponsors 403 00:28:29,260 --> 00:28:33,660 Bangs. I mean, we have hundreds of folks on this project. 404 00:28:33,820 --> 00:28:39,340 We're mapping the human reference map for healthy organs. 405 00:28:40,220 --> 00:28:45,499 Myself, together with upit, we're leading the skin atlas and 406 00:28:45,500 --> 00:28:52,380 creating spatial map of cell types using spatial protein, spatial 407 00:28:52,380 --> 00:28:57,910 proteomics, cell drive, codex, other platforms, similar thing, map 408 00:28:57,910 --> 00:29:01,670 out all the different cell types in our healthy organs, healthy 409 00:29:01,670 --> 00:29:05,870 reference organs, and then also combine that with spatial proteomics, 410 00:29:05,870 --> 00:29:10,230 spatial transcriptomics, as well as single cell signal nuclear 411 00:29:10,470 --> 00:29:15,070 RNA seq. So really multimodal, which I think is where, you know, 412 00:29:15,070 --> 00:29:21,400 if you'd asked us 10 years ago or plus, you know, where, where were 413 00:29:21,400 --> 00:29:26,520 we going? I mean, in terms of building these types of spatial maps. 414 00:29:27,560 --> 00:29:31,240 And you know, that would, we wouldn't have predicted that. So 415 00:29:31,240 --> 00:29:35,040 I think that in and of itself has been very. We can chat more about 416 00:29:35,040 --> 00:29:38,880 that. But that's been super exciting because it's really starting 417 00:29:38,880 --> 00:29:46,400 to show the value of spatial mapping versus dissociated analysis, 418 00:29:46,400 --> 00:29:50,770 which is still extremely important. From, for cataloging and 419 00:29:50,770 --> 00:29:55,570 analyzing all single cells in any given organ. But with the spatial 420 00:29:55,570 --> 00:29:59,730 methods, you're really seeing the different layers of cell types, 421 00:30:00,130 --> 00:30:04,170 the cell interactions, and then combining that with spatial 422 00:30:04,170 --> 00:30:08,050 transcriptomics, then you're really exploding the data. 423 00:30:08,690 --> 00:30:12,930 This sounds like a gargantuan program. Tell me a little bit, how 424 00:30:12,930 --> 00:30:14,610 did you become involved in this program? 425 00:30:15,600 --> 00:30:22,160 Well, you know, like any rfa, the call came out, it, I believe, 426 00:30:22,560 --> 00:30:30,240 was first approved back in 2016, 2017, time frame. And then 427 00:30:30,320 --> 00:30:36,680 we first applied for, with our skin atlas proposal around that time 428 00:30:36,680 --> 00:30:41,800 too, when the RFA came out. And so it's been divided into two 429 00:30:41,800 --> 00:30:46,140 phases. So we. One way is about setting up the data infrastructure. 430 00:30:46,140 --> 00:30:49,380 All the data is being processed centrally, which is also 431 00:30:49,700 --> 00:30:53,940 a huge asset from a community perspective because it's allowing 432 00:30:54,100 --> 00:30:57,420 the data that will be used by the end users. It will be, you know, 433 00:30:57,420 --> 00:31:01,580 we'll have all the provenance and quality control, and it'll be, 434 00:31:01,580 --> 00:31:05,740 you know, have a stamp of a QC stamp on it. And we'll also have 435 00:31:05,740 --> 00:31:09,180 data that's been analyzed by each of the groups themselves. But 436 00:31:09,180 --> 00:31:12,900 this was a massive effort to create that infrastructure, that 437 00:31:12,900 --> 00:31:16,220 machine, if you like. We're now in phase two, which is the production 438 00:31:16,220 --> 00:31:21,260 phase, where we're gearing up, ramping up all the kind of. With 439 00:31:21,260 --> 00:31:24,939 the standardized workflow. So ironed out a lot of the kinks in 440 00:31:24,939 --> 00:31:28,420 phase one also. The other interesting thing about this program 441 00:31:28,420 --> 00:31:32,580 is that not only are we working in 2D, but we're also working 442 00:31:32,580 --> 00:31:37,160 in 3D. And so that's another cool aspect to this where. 443 00:31:37,320 --> 00:31:41,080 So you have multiple sections of the same area so that you're able 444 00:31:41,080 --> 00:31:44,360 to dive. Dive through different planes. 445 00:31:44,600 --> 00:31:49,560 Exactly. So in our hands. So we, yeah, we published our skin atlas 446 00:31:50,360 --> 00:31:54,280 a couple of years ago. It was in 3D. So we did serial sections, 447 00:31:55,000 --> 00:31:59,080 26 serial sections that were. Each section was multiplex. So we 448 00:31:59,080 --> 00:32:03,130 have 14 or so cell types per section, serial section. Those are 449 00:32:03,130 --> 00:32:07,650 registered, and then a 3D volume is created. So that work was 450 00:32:07,650 --> 00:32:12,650 done by Sonia Bose and our team, and he. Then we worked with 451 00:32:12,650 --> 00:32:18,690 Indiana to create the tools to interrogate in 3D the distances between, 452 00:32:18,930 --> 00:32:24,050 say, endothelial cells and nearest. Whatever, you know, your 453 00:32:24,050 --> 00:32:26,650 hub cell is. So in our case, we're looking at endothelial cells 454 00:32:26,650 --> 00:32:30,440 versus immune cells. That was really at the, you know, the sort 455 00:32:30,440 --> 00:32:34,040 of the proof of concept that you could take this data, register 456 00:32:34,040 --> 00:32:39,400 it in 3D and now start to do some analysis. So that was the basis 457 00:32:39,400 --> 00:32:45,000 then for a bunch of tools, including Cell Distance Explorer, 458 00:32:45,000 --> 00:32:50,160 which was part of PubMap, which is now being used across all 459 00:32:50,160 --> 00:32:53,760 other organs in, you know, with different questions about different 460 00:32:53,760 --> 00:32:57,570 cell proximity. And so I think that's the other important thing 461 00:32:57,570 --> 00:33:02,290 about that. I'm happy to jump into any aspect of it. But the 3D 462 00:33:02,290 --> 00:33:08,010 was important because whilst we're used to 2D and we're familiar 463 00:33:08,490 --> 00:33:14,410 with looking at 2D images, organs are 3D, of course. And so 464 00:33:14,410 --> 00:33:21,450 having a 3D reference map is, you know, it's an important step. 465 00:33:21,450 --> 00:33:25,250 It's not and it's not the end step. There's going to be much more 466 00:33:25,250 --> 00:33:28,570 to do in the future. But that was the goal of the program to try 467 00:33:28,570 --> 00:33:31,130 and all to do some degree of 3D. 468 00:33:32,090 --> 00:33:38,610 That sounds incredible. I'm curious just from. With methods like 469 00:33:38,610 --> 00:33:43,530 cell dive, you have to define upfront what markers to look at. 470 00:33:44,250 --> 00:33:47,530 And so how did you go about the process, I guess first of deciding 471 00:33:47,530 --> 00:33:53,030 that you wanted to look at skin versus liver or lung or you 472 00:33:53,030 --> 00:33:57,270 know, pick your organ and. And then two, how did you pick what markers 473 00:33:57,270 --> 00:33:59,550 to look at even within. Within skin? 474 00:34:00,190 --> 00:34:03,910 Yeah, it's a great question. So, so for skin, I mean, skin is 475 00:34:03,910 --> 00:34:08,110 the largest organ in the body. So we are collaborators at yupic. 476 00:34:08,990 --> 00:34:13,070 We've worked with UPIC for a long time and very fortunate to have 477 00:34:13,070 --> 00:34:17,840 great collaborators there and you know, discussing with them, you 478 00:34:17,840 --> 00:34:21,280 know, there was any number of organs we of course could have gone 479 00:34:21,280 --> 00:34:25,880 after, but we decided skin because, you know, skin disease, 480 00:34:26,840 --> 00:34:32,040 inflammation, skin cancer, it's a huge issue that affects everyone 481 00:34:32,760 --> 00:34:36,240 at some point in their life. And it's also very accessible. I 482 00:34:36,240 --> 00:34:40,920 mean, we use samples that in the first few studies that were archived, 483 00:34:41,440 --> 00:34:45,760 but then later we were able to get biopsies, which was really nice, 484 00:34:46,800 --> 00:34:51,840 from donors who provide us with samples from sun exposed and 485 00:34:51,920 --> 00:34:55,960 chronically sun exposed, less unexposed regions younger and older. 486 00:34:55,960 --> 00:35:00,000 So we're able to, we're not only doing an atlas, but we're also, 487 00:35:00,400 --> 00:35:04,480 at least at a small scale, able to compare anatomically within 488 00:35:04,960 --> 00:35:08,840 donors. And so it's sort of, we wanted to bring science into it 489 00:35:08,840 --> 00:35:12,220 as well. That was really came about that whole thinking. You know, 490 00:35:12,860 --> 00:35:21,260 we work with Dr. Luffelo and Dr. Arivarson and John and Ho at 491 00:35:21,260 --> 00:35:25,460 UPIT and they just helped sort of form these ideas around how we 492 00:35:25,460 --> 00:35:30,500 could approach, you know, skin as an atlas with some interesting 493 00:35:30,500 --> 00:35:35,420 questions as well. And then how do we choose the antibody? Well, 494 00:35:35,820 --> 00:35:39,980 you know, the good thing is, I mean when you work with derma pathologists, 495 00:35:39,980 --> 00:35:43,980 they know exactly what biomarkers will mark, which cell 496 00:35:43,980 --> 00:35:48,060 types they'll you know, so we wanted, you know, all the epidermis, 497 00:35:48,060 --> 00:35:53,100 the dermis, the immune cells, the Langerhan cells. So, so we, so 498 00:35:53,100 --> 00:35:57,020 we documented with them what all the likely markers needed to 499 00:35:57,020 --> 00:36:01,980 be. I'm sure we should include from an atlas perspective. And then 500 00:36:02,730 --> 00:36:08,130 separately to this as a sort of parallel strand to HubMap, there's 501 00:36:08,130 --> 00:36:12,650 a. The Human Reference Atlas effort, which is led by Katie Boyner 502 00:36:12,650 --> 00:36:17,890 in Indiana. And they're, you know, creating this monumental map 503 00:36:17,890 --> 00:36:23,850 of all the known cell types and associated biomarkers into one 504 00:36:24,410 --> 00:36:28,800 sort of one master database, if you like. So that again, from 505 00:36:28,960 --> 00:36:33,200 an external. Both from our perspective, could we standardize 506 00:36:33,520 --> 00:36:39,600 which markers we were going to select for the most common cell types, 507 00:36:40,000 --> 00:36:44,159 map that to what's called the anatomical structures and cell types. 508 00:36:44,159 --> 00:36:48,520 Cell types and biomarkers table. That they were documenting 509 00:36:48,520 --> 00:36:50,000 with all the other organs. 510 00:36:50,240 --> 00:36:53,840 Yeah, that's really interesting. And in terms of learning 511 00:36:53,840 --> 00:36:57,870 so far, I know you're still mid straight dream, but what was 512 00:36:57,870 --> 00:37:02,950 the most surprising result for you? And I'm particularly curious 513 00:37:02,950 --> 00:37:10,030 in the homogeneity or heterogeneity aspects, particularly 514 00:37:10,190 --> 00:37:15,110 in three dimensions. It seems like there's potentially some very 515 00:37:15,110 --> 00:37:17,710 surprising results that could result from that. 516 00:37:17,790 --> 00:37:22,600 I think one of the really interesting things which we maybe, 517 00:37:22,920 --> 00:37:27,240 you know, we're used to seeing sort of the classical pictures of 518 00:37:27,400 --> 00:37:32,880 skin in our textbooks and it's very fixed and it always, always 519 00:37:32,880 --> 00:37:36,520 looks the same. But of course, when you're working with samples 520 00:37:36,600 --> 00:37:41,680 from younger, older, sun exposed, less exposed, then you start 521 00:37:41,680 --> 00:37:46,960 to see things like, you know, the skin cell damage and immune cell 522 00:37:46,960 --> 00:37:53,910 infiltration. And you know, in a couple of cases there were. There 523 00:37:53,910 --> 00:37:58,550 was like a minor cyst in the sample and just so, you know, blood 524 00:37:58,550 --> 00:38:04,070 of immune cell. When you come to appreciate how important, of course, 525 00:38:04,070 --> 00:38:09,790 skin is and how well designed it is from the vascular perspective 526 00:38:09,870 --> 00:38:14,470 and fatigue and just, you know, the hair follicles, you know, 527 00:38:14,470 --> 00:38:20,760 the different cell types, stem cells, immune cells. So I think it's 528 00:38:20,760 --> 00:38:25,600 sort of, you know, we probably oversimplify skin, but yeah, there's 529 00:38:25,600 --> 00:38:30,560 20 billion T cells in skin. I mean, we, we caught, you know, maybe 530 00:38:30,560 --> 00:38:31,160 one million. 531 00:38:32,280 --> 00:38:35,240 Wow, that's. That's incredible. That's utterly incredible. 532 00:38:35,400 --> 00:38:39,640 They. So I, I guess I'd like to maybe change gears a little bit 533 00:38:39,640 --> 00:38:45,100 and just talk about with all of these learnings, what you were 534 00:38:45,100 --> 00:38:49,180 doing in hubmap was very much basic science, learning the foundations, 535 00:38:49,660 --> 00:38:54,660 but with these areas of looking at damage and you're transitioning 536 00:38:54,660 --> 00:38:59,340 to thinking about the more clinical side of spatial proteomics. 537 00:38:59,340 --> 00:39:04,860 And so I'd love to hear your thoughts on where spatial proteomics 538 00:39:04,860 --> 00:39:10,860 is going from a clinical perspective. What are the questions 539 00:39:10,860 --> 00:39:14,780 that we're going to be asking with it? What are the places where 540 00:39:14,780 --> 00:39:17,820 this might be part of clinical workups in the future? 541 00:39:18,780 --> 00:39:22,820 I think, you know, what's really stayed the course in terms 542 00:39:22,820 --> 00:39:28,300 of interrogation, maybe clinical value, translational value 543 00:39:28,300 --> 00:39:30,140 is immuno oncology. 544 00:39:30,620 --> 00:39:31,020 And. 545 00:39:33,020 --> 00:39:40,850 Being able to classify or to image different immune cells and 546 00:39:41,010 --> 00:39:45,010 you know, their exhaustion, their functional status. I think 547 00:39:45,010 --> 00:39:50,090 that's still a non, it's still, it's not in the clinic yet. 548 00:39:50,090 --> 00:39:55,770 I mean it will take some time, but it will. To me that's still probably 549 00:39:55,770 --> 00:40:00,010 the, one of the most promising areas as we try to understand why 550 00:40:00,010 --> 00:40:04,490 patients don't respond to immunotherapies. I think it still 551 00:40:04,490 --> 00:40:07,570 takes more time. I mean there's work on going on in simulation 552 00:40:07,810 --> 00:40:14,210 and modeling to understand, you know, you know, why some tumors 553 00:40:14,210 --> 00:40:19,850 are more immunoresponsive than others. You know, what changes that 554 00:40:19,850 --> 00:40:24,890 status over time? Why is, you know, our older versus younger patients 555 00:40:24,890 --> 00:40:28,530 maybe have a different type of so on. So it's still, I think there's 556 00:40:28,530 --> 00:40:32,330 still a lot of research to do. On the other hand, you know, one 557 00:40:32,330 --> 00:40:36,120 of the other, one of the earliest projects, clinical projects 558 00:40:36,120 --> 00:40:41,160 that we were peripherally involved with was with clariant back 559 00:40:41,160 --> 00:40:45,560 in 10 years ago where, 12 years ago actually where pathologists 560 00:40:45,560 --> 00:40:49,360 we work with there felt lymphoma was really a perfect use 561 00:40:49,360 --> 00:40:54,440 case for multiplexing because it can take 20 to 30 different serious 562 00:40:54,440 --> 00:40:58,480 actions to determine what subtype of lymphoma. There's over 563 00:40:58,720 --> 00:41:03,510 30 cells, subtypes of alevo of lymphoma. So, you know, you need 564 00:41:03,510 --> 00:41:07,950 to do a lot of sectioning in order to find out, you know, what 565 00:41:07,950 --> 00:41:11,390 type of lymphoma that is, meaning flow cytometry is possible. 566 00:41:11,550 --> 00:41:16,390 But you still, you know, you might not have enough sample to do 567 00:41:16,390 --> 00:41:19,910 that. So I think, you know, I, and we, you know, at the time there 568 00:41:19,910 --> 00:41:24,390 was a test that was launched for lymphoma, but I feel you might 569 00:41:24,390 --> 00:41:27,290 have been a little early and I think that there's still a need. 570 00:41:27,370 --> 00:41:31,810 We, you know, in the, in the, in the later years, we, it does come 571 00:41:31,810 --> 00:41:36,650 back up again that lymphoma is one that still could be, you know, 572 00:41:36,650 --> 00:41:39,450 very valuable. And the good thing is that there's a lot more 573 00:41:39,690 --> 00:41:41,690 people working in this space now. 574 00:41:42,010 --> 00:41:46,210 So is the thought then that, that in the future it may be that 575 00:41:46,210 --> 00:41:51,730 we instead of doing sort of normal single marker pathology, we 576 00:41:51,730 --> 00:41:54,810 instead might move to a world where we're, we're doing, you know, 577 00:41:55,090 --> 00:42:00,170 full on spatial proteomics and learning much more sophisticated 578 00:42:00,170 --> 00:42:04,530 patterns than just how many cells are expressing X marker. 579 00:42:04,850 --> 00:42:10,330 Yeah, I mean, I think there'll be a few steps still to get there 580 00:42:10,330 --> 00:42:15,970 because it's the relationship between cytotoxic T cell infiltration 581 00:42:16,370 --> 00:42:21,410 and outcomes is still a very powerful predictor. We see it in 582 00:42:21,900 --> 00:42:26,700 all of our studies, you know, even depending on how many patients 583 00:42:26,700 --> 00:42:30,060 you have. But the trend is always there that the higher the 584 00:42:30,140 --> 00:42:34,540 cytotoxic T cells, the better your outcome. But then when you start 585 00:42:34,540 --> 00:42:39,380 to split it down a little bit more and you start to look at, let's 586 00:42:39,380 --> 00:42:42,500 say, the phenotype of those immune cells, you start to see a 587 00:42:42,500 --> 00:42:47,860 little bit more gradation there in terms of, you know, okay, 588 00:42:47,860 --> 00:42:50,460 the patient. There will still be the patients who always do worse, 589 00:42:50,460 --> 00:42:53,700 who have, let's say, cold to tumors. But then I think you can 590 00:42:53,700 --> 00:42:58,380 split it out further with additional information about those 591 00:42:58,380 --> 00:43:02,820 biomarkers, about the functional status of those immune 592 00:43:02,820 --> 00:43:07,220 cells. And also, you know, as we get into, you know, the different 593 00:43:07,220 --> 00:43:12,380 types of immuno oncology therapies, knowing, you know, what 594 00:43:12,380 --> 00:43:16,500 type, you know, the exhaustion status of those, are they functional, 595 00:43:16,500 --> 00:43:21,550 are they able to actually so you could imagine, you know, I don't 596 00:43:21,550 --> 00:43:25,350 know whether it'll be a 30 marker assay, but it probably would 597 00:43:25,350 --> 00:43:28,910 be something closer to maybe 10 at least where you're really, 598 00:43:29,070 --> 00:43:35,150 you need to get sort of, you know, very automated counts. Yeah. 599 00:43:35,150 --> 00:43:40,030 And for sure there's also, and we showed this actually in one of 600 00:43:40,030 --> 00:43:44,820 our papers we published last year on colorectal cancer. The tumor 601 00:43:44,820 --> 00:43:50,700 cells themselves, you know, they have to be responsive to the 602 00:43:50,700 --> 00:43:54,580 effects of the cytotoxic C cells. And if they're what we showed 603 00:43:54,580 --> 00:44:01,380 in a couple of papers, one was both with the World Cultural Surgeons 604 00:44:01,380 --> 00:44:04,220 in Ireland and Queen's University Belfast, we looked at 605 00:44:04,380 --> 00:44:10,650 heterogeneity of apoptosis sensitivity and response to chemotherapy. 606 00:44:10,730 --> 00:44:19,690 So there's sort of this persistent apoptosis resistant cell 607 00:44:19,770 --> 00:44:24,250 cancer cell type that may be not responding to chemotherapy and 608 00:44:24,250 --> 00:44:30,490 may also. And in another paper we show that the proximity of cytotoxic 609 00:44:30,490 --> 00:44:35,890 T cells to their neighboring cancer cells, the neighboring cancer 610 00:44:35,890 --> 00:44:41,200 cells, you can directly quantify the level of Caspase 3 level. 611 00:44:41,280 --> 00:44:46,000 And so, so there's, there's a neighborhood effect. So. So me killing 612 00:44:46,000 --> 00:44:49,480 this cell actually has an effect on its friends and neighbors 613 00:44:49,480 --> 00:44:49,880 as well. 614 00:44:49,880 --> 00:44:53,360 Yeah, and I think that whole space is very fascinating because, 615 00:44:53,600 --> 00:44:57,120 you know, not work we've done, but others have shown you know, the 616 00:44:57,120 --> 00:45:02,280 sort of the, the killing time of a cytotoxic T cell. You know, 617 00:45:02,280 --> 00:45:04,840 this sort of takes three shots. And you know, not just the 618 00:45:04,840 --> 00:45:06,640 proximity even you have to induce. 619 00:45:08,240 --> 00:45:11,160 Well, and I think what you're talking about goes even beyond immuno 620 00:45:11,160 --> 00:45:15,920 oncology, because what you're referring to there's sort of the 621 00:45:15,920 --> 00:45:22,400 historical dogma was hey, I have a kinase inhibitor or I have 622 00:45:22,400 --> 00:45:26,400 a cytotoxic agent, and then it gets to the cancer cell. And the 623 00:45:26,400 --> 00:45:29,440 cancer cell is just like, I'm done, I'm out, goodbye. And it dies 624 00:45:30,160 --> 00:45:34,240 versus there being a necessary interaction with the immune system, 625 00:45:35,050 --> 00:45:39,290 even for drugs that are acting on the cancer cell. 626 00:45:39,690 --> 00:45:43,610 Right, Exactly. Yeah. The other very interesting thing, I think 627 00:45:43,610 --> 00:45:47,330 is that a lot of these studies that we and others have done, these 628 00:45:47,330 --> 00:45:50,970 are adjuvantly treated patients. So the cancer is actually 629 00:45:51,770 --> 00:45:56,290 removed, or at least the colorectal cancer. It's the residual 630 00:45:56,290 --> 00:46:00,570 cells that we're making a prediction or that we're making predictions. 631 00:46:00,570 --> 00:46:03,720 On a tumor that was, that's not there anymore. 632 00:46:04,280 --> 00:46:09,440 So, you know, the sort of very interesting thought that invasion 633 00:46:09,440 --> 00:46:13,240 has already occurred, but still we can make some level of prediction. 634 00:46:13,640 --> 00:46:17,279 And maybe that's because in the end, maybe the tumors are more 635 00:46:17,279 --> 00:46:20,760 uniform from the perspective of, you know, what's left behind. 636 00:46:21,960 --> 00:46:24,560 That's an interesting deep question because you're, you're wrestling 637 00:46:24,560 --> 00:46:28,040 with the reality that the microenvironment and the metastatic 638 00:46:28,040 --> 00:46:32,070 niche is very different than the microenvironment in the primary 639 00:46:32,070 --> 00:46:32,630 tumor. 640 00:46:32,870 --> 00:46:33,830 Right. Yeah. 641 00:46:33,910 --> 00:46:38,550 And yet you still have this cell endogenous characteristic and 642 00:46:39,590 --> 00:46:44,150 how much is it influenced by its friends and neighbors and when. 643 00:46:44,630 --> 00:46:48,630 Yeah. And how much training has the immune system already received? 644 00:46:48,790 --> 00:46:52,510 There has to have been training that is long lasting. It's 645 00:46:52,510 --> 00:46:55,220 a fascinating. I don't know how to solve it. So it's like. 646 00:46:55,380 --> 00:46:58,580 Well, that, that's actually, that's a great segue. So my, my next 647 00:46:58,580 --> 00:47:02,900 question for you was going to be about what can't we do yet? So 648 00:47:02,900 --> 00:47:07,220 the technology has advanced incredibly over the last 20 years, 649 00:47:07,940 --> 00:47:11,460 but there's still things that we might like to be able to do that 650 00:47:11,460 --> 00:47:16,020 we can't do today. So where are the limits of today's technology 651 00:47:16,500 --> 00:47:19,780 that you like? Okay, 20 years from now we should be able to do 652 00:47:19,780 --> 00:47:23,000 X. And today it's a problem, I think. 653 00:47:23,000 --> 00:47:26,920 Speed, throughput, you know, those are, and cost. I mean, they're, 654 00:47:26,920 --> 00:47:31,480 they're, they're not the most exciting things to solve, but those 655 00:47:31,480 --> 00:47:34,040 are the ones that, those are, those. 656 00:47:34,040 --> 00:47:37,720 Are the particular pain Points that, that when you think about the 657 00:47:37,720 --> 00:47:41,680 technology achieving ubiquity, those are the, those are the big 658 00:47:41,680 --> 00:47:42,560 barriers today. 659 00:47:42,640 --> 00:47:46,800 I think, you know, you're competing today against single marker 660 00:47:47,200 --> 00:47:49,920 and you have to, you know, you have to prove out the values and 661 00:47:49,920 --> 00:47:53,560 we've seen scenarios where you know, you're willing to wait as a, 662 00:47:53,560 --> 00:47:55,840 you know, if you're getting good information and it's getting, 663 00:47:55,840 --> 00:47:58,640 you're getting actionable information, sequencing is, you know, 664 00:47:58,640 --> 00:48:04,960 still takes time to do a gene panel. So you know, it would be down 665 00:48:04,960 --> 00:48:08,280 to the value of the data that you're getting as well. So 20 years 666 00:48:08,280 --> 00:48:13,040 from now, I mean it would be amazing if it was as common to do 667 00:48:14,030 --> 00:48:18,710 multiplex imaging as it is to do flow cytometry or sequencing that 668 00:48:18,710 --> 00:48:21,750 it's just built into the workflow. I think we would need to 669 00:48:21,750 --> 00:48:25,350 have, you know, you know, a sort of, at least a sort of a 24 670 00:48:25,350 --> 00:48:30,630 hour turnaround. I think we, things like HubMap, HTAN, Human Tumor 671 00:48:30,630 --> 00:48:36,590 Atlas network and these big efforts will to some extent normalize 672 00:48:36,590 --> 00:48:40,280 how we think about these new technologies. These are going to 673 00:48:40,280 --> 00:48:43,080 be massive amounts of data that are going to be available for 674 00:48:44,280 --> 00:48:47,160 further research. And I think that's a tipping point too because 675 00:48:47,640 --> 00:48:49,800 if you look at cancer genome atlas. 676 00:48:51,160 --> 00:48:54,800 That's right, the value of that has been exponentiated because 677 00:48:54,800 --> 00:48:59,880 it's been analyzed and reanalyzed and reanalyzed a billion 678 00:48:59,880 --> 00:49:04,600 times by so many different researchers and brought in to supplement 679 00:49:05,490 --> 00:49:07,490 so many other parallel studies. 680 00:49:07,890 --> 00:49:11,890 Exactly. Yeah. And I think you know, that's in a sense what it's 681 00:49:11,890 --> 00:49:15,930 going to take as well to get us to the aha's. Yeah, I don't think 682 00:49:15,930 --> 00:49:19,930 we've even our brains have even got there yet. But I can see, 683 00:49:19,930 --> 00:49:23,130 I mean there's you know, heterogeneity types of tumor cells 684 00:49:23,130 --> 00:49:25,970 themselves, the different types of immune cells are present 685 00:49:25,970 --> 00:49:30,450 or infiltration location, you know, those are being studied a great 686 00:49:30,980 --> 00:49:35,020 depth I think across you know, many publications starting to come 687 00:49:35,020 --> 00:49:39,780 out now as well as sort of like that more basic level of research 688 00:49:39,780 --> 00:49:46,420 of understanding, you know, what the transitions from early stage 689 00:49:46,500 --> 00:49:51,180 cancer to aggression, you know, they're having these types 690 00:49:51,180 --> 00:49:54,740 of spatial insights will give us new idea, other new ideas that 691 00:49:54,740 --> 00:49:55,860 we haven't been thought of. 692 00:49:56,020 --> 00:50:00,110 So on that topic. So my last question for you today, you mentioned 693 00:50:00,110 --> 00:50:06,030 soumya earlier and this of course the topic du jour is AI and 694 00:50:06,190 --> 00:50:09,790 how this interacts. And so I'm curious to hear your perspectives 695 00:50:10,350 --> 00:50:13,510 as we start building these libraries, as we start building these 696 00:50:13,510 --> 00:50:19,470 resources as we start looking not just in 2D but in 3D. What are 697 00:50:19,470 --> 00:50:23,750 you already seeing in terms of AI tools coming to bear and what 698 00:50:23,750 --> 00:50:26,440 do you envision for, for AI in the future? 699 00:50:27,400 --> 00:50:32,480 Yeah, I mean, somebody goes, he's the right person to talk about 700 00:50:32,480 --> 00:50:37,200 this. He's one of our team members and has always been the head 701 00:50:37,200 --> 00:50:41,240 of the curve on how we can analyze this data with the latest 702 00:50:41,320 --> 00:50:46,000 AI methods. I mean, some of the stuff that we've done in the 703 00:50:46,000 --> 00:50:50,840 last few years have include the GAN approaches, Generative Adversarial 704 00:50:50,840 --> 00:50:55,530 Network where using that to simulate more data beyond what we 705 00:50:55,530 --> 00:50:59,370 even have. But I think it's the general idea, you know, and we 706 00:50:59,370 --> 00:51:05,370 published that with Emory a couple years ago, looking at H and 707 00:51:05,370 --> 00:51:11,330 E images and creating additional patches derived. So you, 708 00:51:11,330 --> 00:51:15,650 you start with N number of samples and you know, let's say 20 709 00:51:15,730 --> 00:51:20,810 or 100 samples. But then you can start to identify which regions 710 00:51:20,810 --> 00:51:24,250 within the sample are associated with aggressive disease 711 00:51:24,730 --> 00:51:29,130 and which are, let's say more innocuous regions. So you're trying 712 00:51:29,130 --> 00:51:33,850 to essentially balance your data set by creating more of these 713 00:51:33,850 --> 00:51:37,850 aggressive patches. And you know, one of the problems with these 714 00:51:37,850 --> 00:51:43,570 types of analysis, predictive analysis, is imbalanced a very small 715 00:51:43,570 --> 00:51:47,230 amount of aggressive regions and then you have a lot of normal. 716 00:51:47,870 --> 00:51:50,750 So I think it's sort of fascinating to think that we can, 717 00:51:50,750 --> 00:51:55,070 you know, create more data. You know, you sort of wonder, okay, 718 00:51:55,470 --> 00:51:59,230 is it, it's real? Because it's based off of something that was real. 719 00:51:59,470 --> 00:52:02,990 But you know, how far can you go? And I, you know, that's, I guess 720 00:52:02,990 --> 00:52:05,150 a general question for the field, but. 721 00:52:05,310 --> 00:52:08,990 It'S a deep question because on it, it was actually quite a surprising 722 00:52:09,070 --> 00:52:14,520 result that you could, could add noise, rotate, do these very 723 00:52:14,520 --> 00:52:18,440 simple transforms and the AI methods would actually learn better. 724 00:52:19,240 --> 00:52:22,640 And as a human we're like, well, it's the same image, it's just 725 00:52:22,640 --> 00:52:26,440 rotated 90 degrees. Why should the AI do better? And yet they did. 726 00:52:26,840 --> 00:52:27,960 It was astonishing. 727 00:52:28,360 --> 00:52:31,760 Yeah, no, it is, it is astonishing and as well, because 728 00:52:31,760 --> 00:52:35,000 these technologies are expensive. So can you create more 729 00:52:35,000 --> 00:52:41,090 data, you know, without having to spend a lot of money? But you 730 00:52:41,090 --> 00:52:45,250 know, the question is how diverse can you make your data sets? 731 00:52:45,250 --> 00:52:49,650 And you know, are rare cell types, they may not be captured. 732 00:52:49,730 --> 00:52:53,090 So I think it's, it's sort of got its place depending on what your 733 00:52:53,090 --> 00:52:56,730 question is. And then I think also, you know, the deep learning 734 00:52:56,730 --> 00:53:04,010 methods for cell analysis and you know, features within tumors 735 00:53:04,010 --> 00:53:07,250 using the H and E, that's another in a whole other area. But 736 00:53:07,250 --> 00:53:11,210 you know, there's A lot within the H and E image that we, you know, 737 00:53:11,530 --> 00:53:16,330 is untapped as well. So that could also be, you know, very nice 738 00:53:16,330 --> 00:53:19,850 hybrid. But yeah, I think AI can help on all those ones. Also 739 00:53:19,850 --> 00:53:25,010 the large language models in trying to just distill down, you 740 00:53:25,010 --> 00:53:28,810 know, what information, actions, you know, what does it all 741 00:53:28,810 --> 00:53:29,130 mean? 742 00:53:29,710 --> 00:53:31,790 That makes a lot of sense. So I'm going to put you on the spot 743 00:53:32,110 --> 00:53:36,510 with my last, last question which is really again a little bit 744 00:53:36,510 --> 00:53:42,230 of a forward looking. If you think about yourself 10 years from 745 00:53:42,230 --> 00:53:46,910 now, 15 years from now and you think about, you're looking back, 746 00:53:48,030 --> 00:53:52,070 what have you learned? So you've had all of this time, you've 747 00:53:52,070 --> 00:53:55,150 had all this infrastructure, you've done so many of these big 748 00:53:55,150 --> 00:54:00,260 projects. What is the, what is the, that exciting thing that you've 749 00:54:00,260 --> 00:54:02,140 learned over the course of this journey? 750 00:54:03,100 --> 00:54:09,740 Probably the immune response is really something very remarkable, 751 00:54:10,060 --> 00:54:15,020 not surprising, but how much you know, I think there's so I hope 752 00:54:15,100 --> 00:54:19,660 in the next 10 years that that's even more solved than it is 753 00:54:19,660 --> 00:54:23,340 today. But just when I look back, you know, over the last 10 754 00:54:23,340 --> 00:54:26,900 years when it was even maybe more than last 10 years, maybe that's 755 00:54:26,900 --> 00:54:31,020 15 years ago, it, we did not even analyze the insights. 756 00:54:31,180 --> 00:54:33,060 You just ignored them. They were just, they were there, they 757 00:54:33,060 --> 00:54:36,420 were annoying, they were like no, no. 758 00:54:36,420 --> 00:54:40,940 You don't, they don't need to look at those. And then. But I think 759 00:54:40,940 --> 00:54:47,430 just, it's just profound how that's changed and just the, you 760 00:54:47,430 --> 00:54:51,550 know, the variability amongst the different, the functions of the 761 00:54:51,550 --> 00:54:54,110 different types of immune cells, their interaction, their cross 762 00:54:54,110 --> 00:55:00,550 talks, their metabolism, how that plays into vaccines for, for 763 00:55:00,630 --> 00:55:05,790 cancer vaccines. I think, you know, that's still to pan out, but 764 00:55:05,790 --> 00:55:07,430 I hope that's another holy ground. 765 00:55:07,830 --> 00:55:10,390 Anything else that you're, you're excited about as this field 766 00:55:10,390 --> 00:55:11,350 develops and evolves? 767 00:55:12,230 --> 00:55:18,200 Yeah, I think the really wonderful, phenomenal thing is just 768 00:55:18,200 --> 00:55:22,760 all the next generation of scientists, junior investigators 769 00:55:23,240 --> 00:55:27,920 who are now working on this data in totally different ways. You 770 00:55:27,920 --> 00:55:31,520 know, we see this on hopmap with the junior Investigators and 771 00:55:31,520 --> 00:55:37,120 their Jumpstart program and just phenomenal. I mean they're coming 772 00:55:37,120 --> 00:55:45,490 up with amazing ideas around analyzing cells, neighborhoods. You 773 00:55:45,490 --> 00:55:48,810 know, you've got people like John Hickey who's now in Duke and 774 00:55:48,890 --> 00:55:53,250 Chen Shenzhou in Stanford and others and that there's like a list 775 00:55:53,250 --> 00:55:58,450 of folks who just, they get it and they're just thinking in a totally 776 00:55:58,450 --> 00:56:01,610 different way. And I think that's going to change the field 777 00:56:01,930 --> 00:56:05,730 as well. So that's, that's very exciting. I can't wait to see 778 00:56:05,730 --> 00:56:07,830 what they do in the next phase. 779 00:56:08,390 --> 00:56:12,590 That's awesome. Well, thank you. Thank you again. So for our 780 00:56:12,590 --> 00:56:16,710 listeners, you've been sitting here chatting with Fiona Ginti, had 781 00:56:16,710 --> 00:56:20,110 an incredibly fun conversation, learning about the 782 00:56:20,110 --> 00:56:26,550 Hubmap Consortium and the role of spatial proteomics in the clinic 783 00:56:26,550 --> 00:56:33,130 and where it's going, as well as the future involving AI and the 784 00:56:33,280 --> 00:56:37,040 the immune system in cancer as well. So thank you so much, Fiona, 785 00:56:37,040 --> 00:56:37,760 for joining us today. 786 00:56:38,160 --> 00:56:39,840 Absolutely. Thanks, Tarek. 787 00:56:40,320 --> 00:56:42,800 All right, thanks, everyone, for joining us. And I hope to see 788 00:56:42,800 --> 00:56:45,760 you again soon on another episode of Translating Proteomics. 789 00:56:49,199 --> 00:56:52,920 We hope you enjoyed the Translating Proteomics podcast brought 790 00:56:52,920 --> 00:56:56,920 to you by Nautilus Biotechnology. To contact us or for 791 00:56:56,920 --> 00:57:01,360 further information, Please email translating ProteomicsAutilus 792 00:57:01,360 --> 00:57:01,840 bio.