 Good afternoon, everyone. Okay? Yeah, good afternoon, everyone. My name is Jeff Schloss. I'm the director of the Division of Genome Sciences in the Extramural Research Program at NHGRI. Let me welcome you to the continuation of the seminar series, A Quarter Century After the Human Genome Project's Launch, Lessons Beyond the Base Pairs. I'm pleased today to introduce Dr. Marco Mara as the fifth speaker in this NHGRI seminar series commemorating the 25th anniversary of the launch of the Human Genome Project. You may have noticed that Dr. Mara is not physically present, but instead is joining us by video feed. Unfortunately, a last-minute scheduling conflict arose for him, and rather than canceling and thus missing out on Dr. Mara's presentation, we decided it would be better to proceed even if it meant using the wonders of the Internet to bring us together electronically today. Dr. Mara received his Bachelor of Science degree and PhD degrees from Simon Frazier University. He pursued post-doctoral training at Washington University School of Medicine in St. Louis, eventually holding a number of positions at the university's genome sequencing center. There he contributed to the Human Genome Project by developing and implementing technologies to map the human genome while working with Bob Waterston and the team at WashU. In 1999, he moved to the Genome Sequencing Center at British Columbia Cancer Agency and then joined the Department of Medical Genetics, University of British Columbia, in 2000. Dr. Mara now serves as the head of the Genome Sciences Department and the director of Canada's Michael Smith Genome Sciences Center at the British Columbia Cancer Agency. He's also head of the Department of Medical Genetics Faculty of Medicine at the University of British Columbia and Adjunct Professor, Department of Molecular Biology and Biochemistry at Simon Frazier. The Genome Sciences Department focuses on the development and application of genomic, bioinformatic, and proteomic assays to problems in human health, particularly cancers. The Genome Sciences Center represents an internationally acclaimed innovation group that encompasses DNA sequencing, bioinformatics, and technology development in Vancouver, Canada, with a major focus on the study of cancer. Dr. Mara has received many awards. To mention just a few of them, he was appointed as a Fellow of the Royal Society of Canada in 2007 and elected to the Canadian Academy of Health Sciences in 2009. Dr. Mara is a member of the Order of British Columbia and recently he was listed in the 2014 and 2015 world's most influential scientific minds by Thompson Reuters. Dr. Mara's career achievements so far are undeniably impressive. Undoubtedly his contributions to genome science and cancer research will continue to influence our knowledge of human biology and to improve human health. So please join me in welcoming Dr. Marco Mara whose talk is entitled, Let's See if It's the Same. Pretty much. From back clones to cancer genomes, the role of the HGP in launching a career in science. Marco. Well, thank you very much, Jeff, for that very kind introduction. I apologize for not being able to be there with you all today. We are attempting a recruitment and so I'm required to participate in that. I'm very grateful to Chiara and Chris and Jeff and Eric for giving me the opportunity to participate in this somewhat unusual venue. I hope it works for you. I'd be interested in whether that is actually at the end of all of this. So the topic that Eric and I chatted about really was around this business of how the Human Genome Project propelled me forward into a career in science rather than the things that we're doing right now. How did this experience shape me for this career that I have invested the last quarter century and I guess we're very nearly that. So I'm going to spend a fair amount of time talking about the good old days because I think the good old days are fundamental to how I approach problems and how I approached the wonderful opportunities afforded to me in the Human Genome Project working with luminaries like Bob Waterston and John Sallston. So the relevant, hopefully relevant, personal history is shown here on before I go on. I forgot to tell you that I'm not in the room by myself. So I have some very kind students here with me who felt sorry for me sitting in a dark room all by myself and have come to lend their moral support and so they are in the room with me here. Okay, so in terms of the personal history business, so Jeff told you that I did my PhD work in Burnaby, B.C., which is just beside Vancouver, and that was all about really worms and worm genetics and trying to identify and clone out genes that when mutated yielded particularly interesting phenotypes. That experience was absolutely fundamental to my decision then to go on and pursue training at WashU, working with Bob and Rick and Elaine and all those good folks. And it was during that time towards the end of my tenure at WashU that I was exposed to various opportunities in cancer research and we're not going to touch on those today, but those were very important experiences that I want to acknowledge up front and then since 99 of course we've been doing things in the British Columbia, Canadian and international environment where now I find myself musing on how we might use fruit flies to pursue some of the interesting problems that are emerging from our cancer genomics efforts. But in 1989 I was more or less unaware of a thing called the Human Genome Project and it was during that same year that this important paper was published by Jim Watson and Elk Jordan, which laid out a description of the evolving Human Genome Project and I was aware of this paper, although I had absolutely no idea that a few years after it was published I would be in some way or another engaged in the conduct of the program. Instead in 1989 I was confronted with this beautiful organism, Sina-Raptagus elegans and 25 years before I encountered C. Elegans as a research tool Sidney Brenner had published his seminal work on the genetics of the animal and its utility as a genetic model organism and so this this was a holy grail to S. Elegans geneticists back in the day this work of Brenner and then following on that the work of Salston and Horvitz in mapping out the gene lineage or the cell lineage of the animal and that's trying to get my pointer up here so you can see where we're at, well there we are so that's what's depicted here is a post embryonic cell lineage of the worm and then of course a number of people were continuing to do intensive mutagenesis screens looking for genes of particular interest using chemical mutagens, radiation, things of that sort and that's basically what our lab was engaged in at the time. So the way that one approached one's studies back in those days really was well you know identify some interesting phenotypes and then off you go go clone the gene and prove the thing that you had cloned was in fact the entity responsible for the phenotype when it was mutant and so this cloning business was incredibly laborious even in a genome as small as the worm about a hundred million base pairs we now know and so meandering around in the genome trying to find the base change that identified the mutated function was was exceptionally tedious and would have been much more tedious had it not been for this wonderful physical map and so Coolson and Salston, Brenner and Carnes published paper in PNAS in 1986 which laid out the evolution of a clone-based physical map of C elegans and so by the time I entered into my PhD studies this clone map had been expanded and was growing and well it wasn't complete was already making an enormous impact on how people approach gene cloning. Waterston then got involved along with Yuji Kohara and they brought in another very important reagent into the mix and these were yaks and so what's shown below here are some short lines these are generally cosmic clones and then these longer lines these depict yeast artificial chromosome clones and these reagents these clones representing the genome were positioned one against each other by virtue of a couple of features and one was this this fingerprinting method that had been devised by Allen and John in which clones were subjected to restriction enzyme digestion and the restriction fragments were labeled and then resolved on polyacrylamide gels and so what this shows is a marker and then lanes of clones and then another marker and lanes of clones and so on and so forth and so the idea was that each clone would yield a stereotypical pattern of restriction fragments and commonalities between restriction fragments could be used to infer that two clones were at least partially overlapping and by doing this over and over and over again in a random way one would build up a deep map of the genome and this deep map then would have these clones arranged with respect to each other now where it became even more important from the cloning perspective is I think shown quite well on this particular slide so what we have on the top here is a genetic map of the worm and and these markers generally have been positioned with respect to each other on the genetic map using standard genetic crosses mapping against other mutations and also against rearrangements such as deficiencies and so the position of these genetic entities on the map was known and then you could use things like the breakpoints of these deficiencies and map those onto cosmic clones and thus define the interval in which the gene of interest must line in this case this is a paper from the Sternberg lab in which the cloning of lethal 23 which turns out to be the worm homologue of the eG at the human eGFR receptor was reported and so this very quickly was able this genetic mapping followed by use of the physical map was able to get you to quite a small interval and then you could test each one of these underlying cosmic clones shown here for their ability to rescue the phenotype of interest when injected into the germline of the animal so so this was amazing because through the availability of the clone resource and through the very extensive genetic tools available to see elegans researchers one could localize the position of the gene one had a clone 40 kilobase clone in case of a cosmic or much larger ones in case of yaks and and these reagents could be obtained so you could phone or email they have email even back in those days and you can email to to request the clone and Alan Coulson would put the clone in an auger plug and mail it to you and it would show up and you could you could work with that as a cloning reagent so this was amazing and so we in the lab that I was working in many many see elegans labs around the world our research was transformed by the availability of this map and so these people who were involved in launching the whole thing became you know my scientific heroes at the time so there's Bob Waterston up there and John Sulston here Alan Coulson Sydney Brenner John here or sorry Alan is standing beside how people used to interrogate the map in those days this is a series of pages printed out and stapled to the wall at Cold Spring Harbor where the map is displayed and people would wander along and ask Alan well you know how's the content coming in such and such an area and so on and so forth and and this was really a revolutionary thing so those people were were viewed as as the pioneers and and I was quite keen to learn more about how such people thought and how they conducted their science they were not the only people that had contributed to the map however this business of Yaks I've alluded to as being very very critical and so what the Yaks did was allow longer stretches of the genome to be linked and so they would span regions that were unclonable and in some regions that were unclonable and Cosmets and overall contributed the continuity of the map so Yaks were incredibly important and in a paper in 1987 by Burke et al Yaks are described so so Maynard Olsen became a very important figure even in the worm community although it wasn't necessarily well recognized what the contribution was in the early days the acts became fundamental to closure of the map but that's not the only thing that Maynard contributed to my early thinking he published this very nice paper in the same issue as Coulson and Solston published their paper on a rather different approach for genomic mapping this in yeast now where smaller insert clones were used but a fundamentally different method for producing the fingerprint was employed and so shown here is not a polyacrylamide gel but rather an agarose gel very very high resolution very carefully cast and run DNA fragments visualized through the application of ethidium bromide and so these were were representative in some in a lane here of fragments from a clone these were representative of the total insert of the clone being analyzed in the Solston and Coulson method the total clone insert was not visualized but rather only a portion of it so this gave us the opportunity when you summed up all these fragments to to know what the clone insert size was and to know the sizes of the individual restriction fragments themselves and these could be arranged to produce these these very wonderful high resolution and deeply redundant restriction maps that span large regions of the yeast genome so at the time this business of making maps as a way to access the genome was being driven by by Maynard and by John and Bob and Alan and around about the time that I was finishing up my my PhD I thought well you know what is it that that I must pursue what am I keenly interested in and this business of the genome by this time it had infected me I needed to know how to interrogate the genome and I needed to see large stretches of these maps to be happy and I needed to see sequence and so I went to wash you and interviewed with with Bob Waterston and Rick Wilson in 1993 in October and what I came to do was sell them on the idea that perhaps what we ought to be doing as a companion project to the C. elegans project was mapping and sequencing a nematode that was related to C. elegans this this one called C. brixie senorabditis brixie and the rationale was as follows so if if and we knew this from various studies that had been published where short snippets of sequence had been investigated if you did a sequence alignment of C. elegans and C. brixie C. e. and C. b. here what you could see is a substantial areas of sequence conservation and that's indicated for example in all these little dots that's conserved sequence and that these substantial regions of sequence conservation tended to be found in regions that were either coding or up here regulatory in nature and if you were to look at an intron down here substantial sequence divergence could be seen so this became a way to find genes and the elements that that control the genes and so the idea was sequence C. elegans sequence brixie align the sequences and and use the resulting orthology or hermology if you like to identify coding sequences in the regions that regulate them and so that's what I what I really wanted to do as for my postdoctoral work and that seemed interesting and I now know that these thoughts were not unique in any way people had been thinking about doing C. brixie for some time I went back and wrote my thesis and thought about where I was going in life and all these kinds of things and about eight months later showed up in August of 1994 I think it was to begin my work and was very pleased to make some friendships that I think have lasted for a quarter of a century and hopefully beyond these are people that were absolutely fundamental in bringing me into the genome center and providing all manner of opportunity a very green kid from the boonies and introducing that very green kid to to the wonders of the genome so Rick and Elaine very influential Ladina Hillier in the informatics space an amazing talent and Bob and Lucinda Folsom were very welcoming when I arrived and did their very best to teach me things with with maybe some success I don't know so other additional important figures in in those those early days are shown here so Stephanie Chisseau right here who's now with laxosmith Glein Mark Johnson who's ahead of genetics in Denver I believe these folks were in the same office as I was put and and we established a close working relationship Mark was busy annotating yeast sequence he was you know allegedly on sabbatical playing with these sequences and Stephanie was coming in as a postdoctoral fellow from Bruce Rose lab in Oklahoma and Stephanie knew all the process you know Bruce Rose lab had been doing this kind of work she was a sequencing expert I was not a sequencing expert I was a C. elegans guy and so Stephanie did her very very best to take me under her wing and exposed me to the mysteries of the process and and the tools being used Mark made made all manner of rather different impressions upon me he introduced Stephanie and I to the wonders of white castle and we frequently ate there for reasons that I simply can't defend anymore he also introduced to me various secrets for avoiding administrative entanglements and the colorful language he used can be repeated here but we had a lovely interaction and it was during these these early days that I met this fellow who is well known to you all that's Eric having a conversation with Maynard taken from a photograph that I think appeared in nature so these were the people that more than any other single thing or or collection of tools made the experience at wash you so incredibly wonderful and as I arrived and Stephanie was teaching me about how you sequence genomes and mark was teaching me about white castle I was exposed to to the process of sequencing and this process was was reported in a paper in 2005 and by this time it gone through many iterations but I think it it serves the purpose for which I intend which is to lay out this this pipeline which was really built around processing individual large insert clones and for the most part and see elegans these these were cosmos and so there was a group that was involved in mapping the clones and there was a group involved in manufacturing sequencing libraries from the clones and then the clone the clones carried in in bacteria would be plated and and picked into micro tighter plates using robots and so on and so forth and then at this process the clone sequences went on to this this group called the finishing core and the finishing guys were people like Bob and Lucinda and others and and they understood how to take their collection of sequences that emerged from the sequencing machines and assemble them and achieve a very high quality sequence product that was that was contiguous and so it was decided through some kind of a discussion that occurred between between Stephanie and Bob and Lucinda I'm sure that I ought to learn how to how to do this finishing business and and that would be a good thing for me to wrap my head around working out the topology of the relationships of these these sequence contigs and stitching together a final finished product and through that I would learn about sequencing so so Bob Fulton decided that I really had to have my own clone to finish and he gave me this clone M03A1 now M03A1 is as you can see up here on chromosome 2 of the worm back in those days M03A1 had only a couple of neighbors it had a cosmic over here and had a cosmic over here and so to span this region of DNA sequence in the worm genome it really was M03A1 or nothing and because the ambition was to produce as finished a product as possible this this clone needed to be finished and so my task was to to finish it and I'm unhappy to report that I failed miserably in that task so the objective was to produce basically a bunch of gene predictions from the sequence of that clone and as you can see in this this nice GenBank entry today that's all been done but at the time I confronted it the clone was in all little pieces so here's a piece of the clone and here's another little piece and another little piece and another little piece and another little piece they're all these sequence pieces and what this picture shows this is a program written by Jeremy Parsons who is an informatician at WashU at the time the lines connect repeated sequences so there were all these sequence repeats some of them arranged less some of them spanning contigs some of them quite complicated it was these sequence repeats that that defeated me I couldn't make heads or tails out of this and so I was a miserable failure I am happy to report however that in 2012 there was an update to the entry and somebody actually finally did finish this thing so it's it's not a gap with my name on it in the genome any more and of course we all know that the sequencing effort was was largely successful so I want to go back to that that slide of HDB and I just want to make a couple of points here so there's MO3A1 that at the time was all lonely in a gap this this clone materialized later as did many others and and they're shown here so these other clones now are are showing up due to efforts in Vancouver a fellow named Don Mormon played a role in mapping these Cosmets or Fosmets rather the Fosmets now are were different than Cosmets they had an F factor origin of replication they propagated a single copy the Cosmets were multi-copy they were much more stable in in bacteria and so the point I think then the take home message that I got out of all of this is that while you know having a bunch of different vectors was absolutely essential if you were going to have a high quality map you needed things that complemented each other some grown in yeast some grown in bacteria you needed single origin multiple origin you needed all these these different clones integrated into an effective map a real breakthrough in in mapping and and cloning very large chunks and being able to propagate them in bacteria came with the report of of Bax which were capable of carrying 300 kb of DNA and and had an F factor origin of replication which could maintain clones and single copy and so with the advent of Bax which were reported in 92 here but which were in use before then with the advent of Bax and the realization that clone maps could absolutely be used to fuel sequencing also the realization that I ought not to be a finisher this led really to a new assignment and that assignment was really you know the problem of how to build a clone map to organize the human genome sequencing so this this problem was presented to me at a time when a decision had been made that maybe the most important thing to do wasn't to pursue this he brings a genome anymore but due to various occurrences that perhaps the human genome ought to be the target and the rationale for making a clone map to to organize human genome sequencing I recall as being something along the lines of what's shown on this particular slide which was basically that these large insert low-copy clones could represent with reasonable fidelity large portions of the human genome and because they were clone based could be integrated with the the other maps that were available for human SDS maps probe based maps so on and so forth and this would enable contigs assembled from these large insert clones to be placed on to answer the human genome now it was it was pretty clear that this map needed to be a fairly significant redundancy that's what the yeast and worm examples had told us you could with a deeply redundant map you can achieve contiguity you would have some estimates to the clone overlap and if you sequenced overlapping clones the idea would be that you would have over you would have long reaches continue contiguous sequence in the event that a clone was not of high fidelity or perhaps was not obtainable or been lost there were replacement clones made it could be made available as a consequence of this deep redundancy and then finally having the sequencing effort organized by clones was I think a particular advantage reviewed as being a particular advantage in the case of the public project because clones could be allocated to different sequencing centers in essence allocating chunks of the genome to these various centers they could work in parallel with minimal wasteful redundancy of effort and this would speed the production of the human genome sequence and so I was privileged to be to be offered the opportunity to think about how such a map could be built and it's at this point that I want to make some other important acknowledgements to folks that provided all kinds of assistance and support and enabled the creation of a human clone map Tami Kuchaba was a person who worked with me through thick and thin for five and a half years helping with the fingerprinting effort and the express sequence tag effort I'm going to talk about that Jeff Wessner and John McPherson good friends who commiserated as appropriate when times of difficulty emerged Peter of course was was very important to me and made excellent clone libraries Kerry Soderland had developed a software package working with John and Bob at the Sanger Institute called FPC and FPC turned out to be available ready to use could be adapted for the kind of clone fingerprinting that we eventually decided on and and she was a big supporter of moving the technology and helping us with her code and then finally Dan Furman who was an engineer at Wash U electrical engineer I recall who got very interested in the problem of automated detection of restriction fragments through the production of software that we collaborated on called band leader so those these were some of the folks that maybe you don't hear about all that often and in the development of the human map and the resulting sequence but they were really critical to the effort so we we had to imagine how we were going to do this and of course there were a few options available immediately we could try and replicate the method that Alan Coulson and and John Solston were expert at and so that was taking these clones and labeling restriction fragment ends and then running a representation of the restriction digest of a clone on a polyacrylamide gel and then there was a Maynard's approach which involved purification of DNA rather simpler restriction enzyme digestion and resolution of the fragments on an augurus gel and after experimenting with these and a few other ideas and considering the magnitude of the problem that that many hundreds of thousands of clones would have to be fingerprinted and that these were large clones we we eventually concluded that the way that we were going to do this was essentially using augurus gels and and mini preps to to process clones and this was featured on on the cover of genome research in 1997 those of you that have thought a little bit about augurus gels and have run them know that that this is not a standard depiction with the top of the gel being down here in the bottom being up here but I guess it appealed to the artistic talents of genome research and so that's the way it's presented as well as our rationale here for pursuing this particular approach and so the strategy that we reported on in this particular paper was as follows so we would take probes or sts's and we would use these to identify arrayed clones so these would be clones that have been spotted on to on to hybridization filters for example or subjected to PCR assays these clones then could be fingerprinted and then the fingerprints could be compared the idea being that a specific sts would identify in a deep enough library clones that overlap and you would like to know more about the identity the nature of those overlaps where the clones completely identical where they slightly different that one extend further to the left and the other extend further to the right and so on and so forth and so that was the general strategy that we pursued and you would produce data using a lot of technology development that went on to do this these kinds of high-resolution gels where very tiny amounts of DNA prepared in 96 well format by this time we're digested and resolved by arborist gel electrophoresis and visualized by cyber green and so what you can't tell easily is this this band here which is up in 20 kb range all the way down to a few hundred base pairs down here so that was the approach that that we reported on and we went on in the paper to show that that clones identified by common sts's could in fact differ and that you could see new bands arising as you walked across a contig and if you did this deeply you could extend the contig recognizing the overlaps finding the new fragments and using these new fragments to walk further distally we also showed in the paper that the restriction fragments were were quite accurate and so within a few percent of real size except for clones up here in the 12 kb range or greater these fragments could not be accurately sized to much better than about three to five percent so that was the basic technology and that that we eventually evolved and then applied and reported on in this paper which was published in the same issue as the initial sequencing and analysis and what you can see here now is how an evolution of the gel format now we have a hundred and twenty one lanes of data by this time we're not using sts's anymore we're just doing mass fingerprinting in a random nature in a sense shotgun fingerprinting all the clones and libraries assembling the fingerprint patterns to each other and then as a consequence of this assembly having these very very large contigs of back clones for the most part which could then be placed onto sts maps by hybridization or actually by sequence data directly so that became the way that that we contributed in fingerprinting map to the sequence effort but by this time by the time this paper came out we were already in operation here and so although I'm listed as participating from the wash you side we had set up shop here and Stephen Jones the head of informatics that are at our center had devised a way to parallelize FPC so that very rapid of old genome assemblies of this sort could be completed and so that was a contribution made to the overall effort from Vancouver and of course we all know that the secrets reported here was was generated largely from from these kinds of clones in a consortium of a sort that I have never before never seen it was really quite amazing you'll be happy to know though that that that you know I had not forgot about sea brexit although I hadn't been engaged in it so somewhat parenthetically the genome did get mapped and it did get sequenced and this was reported long after I had departed from wash you so it did happen and remains a valuable comparative data set to this day so round about 1999 or so I had the opportunity to consider a role at the BC cancer agency and so I was approached very believe I dr. Michael Smith who is shown in this particular photograph he was a laureate in 93 for site direct immunogenesis and he had been approached by guy named Victor Ling who was the vice president researches cancer agency and Victor had a vision in which he would he would very much aim to set up a genome center focused on cancer and he called Mike up Mike at the time was pursuing basically a research sabbatical although Mike called it a postdoc with Maynard Olson in Seattle and so Mike and Maynard were friends and Mike was having a great time working with Maynard on various problems related to the genome and but Mike was persuaded to to become the founding director recruited myself and Steve Jones reuniting us in essence we had been graduate students together in in Burnaby and he had gone to the Sanger Institute to work on genome sequencing from the bioinformatics side and I had gone to watch you and so we were collaborators over this period of time so a decision was made and and I along with Steve joined Mike and the first thing that we really got up to was was more of his fingerprinting business more of his clone mapping and so this became possible because thanks to Bob Waterston and the National Human Genome Research Institute we were able to enter into a subcontractual arrangement in which we started to to fingerprint clones to support the mouse effort at least initially and so a variety of technology development exercises transpired over the years from loading devices capable of loading 96 walls at a time that's what's shown here with fancy ferromagnetic fillings we constructed special fridges to keep our beautiful agarose gels at the proper humidity we had industrial scale agarose gel pouring big huge devices and cooling implements that we would deploy and and then staining cabinets where we could stain many many many of these these gels using cybergreen and prior to imaging them and so we eventually ended up with a setup this is one bench there were four like it double-deckers you see these gels running each gel is actually two gels so there's 242 lanes of data on each one of these things and these things would chug away night and day and make an awful racket with these peristalsic pumps in operation people would walk around with earphones so to block the noise but the data were quite amazing and by this time the band leader software that Dr. Furman had developed with us was being used in a mostly automated way with 97% sensitivity and 95% specificity so most of the fragments even those in multiplets such as these guys here were correctly identified and this then became the foundation for the eventual production here of something like 5.5 million fingerprints mostly backs for 63 different species and much of this was funded by the National Human Genome Research Institute and so that those efforts contributed to most genome maps and there's the Norway rat for example macaque was part of it cattle part of it and locally while we were contributing all those data we started to develop reagents for micro rays back bone micro rays Martin Shrevinsky used the map to identify a set of back bones about 30,000 clones that span the genome and this was then deployed to generate DNA micro rays looking at copy number changes in cancers along the way Martin decided that you know it was absolutely critical that we have some way of visualizing these clones and the relationship to each other and so we invented this thing called surcos and I think most of you have probably seen depictions of this it's now used not only for visualizing back clone fingerprints actually it's not used for that as far as I can tell but rather visualizing sequence relationships it has become a very common way for people to display sequence data and that was done by Martin Shrevinsky so back clone mapping became the dominant high throughput large-scale production activity at the genome center from you know 99 until about 2006 and so for a seven-year period that was primarily what we concerned ourselves with believing that the production of these data was contributing usefully to to high quality reference genome sequences we were also doing some some modest sequencing of course and again I had the opportunity to collaborate with Eric Green in a variety of papers on the mammalian gene collection Bob Strasberg as well we had a brief interlude with the SARS virus in 2003 and mostly got actually quite quite good pardon the expression at making DNA sequencing libraries from limiting amounts of material including micro dissected material and laser captured dissection collected material and so this this proficiency with libraries and the large number of very diverse libraries and we were making largely for purposes such as SAGE and EST sequencing led directly to the invitation from Selexon in 2006 to participate in a technology development exercise in which our particular role was going to be developing methods for library construction suitable for analysis on this absolutely revolutionary brand-new platform so that was that was a milestone moment in our evolution as it was for many others and signal the departure away from back on fingerprinting as a dominant activity it was shortly after this point in time when we retooled and focused our energy on this then-new next generation DNA sequencing and we're fortunate to to be involved in a variety of different consortia over the years in which this this sequencing was deployed in support of various things so the the center you know has grown as a direct consequence of all of that history to today occupy 50,000 square feet across two sites and employees about 320 and has a very training emphasis having trained more than 500 undergraduate graduate postdocs over the years today we we are heavily invested in next generation sequencing technology as well as mass spectrometry infrastructure and have a fairly considerable compute infrastructure and have used this to to focus mainly on the lignancies and over the course of 2009 would be down here some place have generated on this particular trajectory now slightly more than a petabase of DNA sequence from something like 63,000 individual libraries so that's the the business of clones to cancer if you like that I alluded to in my second slide what I'd like to do now in the next few minutes is is just talk a little bit about malignancy I couldn't resist the opportunity to make a few observations and these I think are to be considered in the context of where I think we're going with our own effort here and where I believe others may be going with the application of genomics tools to characterize cancers so in this particular milieu this big high chart of things is micro RNA seek and so there's been many many thousands of micro RNA seek libraries that we've generated and processed and contributed to the TCGA over the years as well as some some RNA seek and of course we all know that that ICGC TCGA these entities have produced if you like reference sequences for for many malignancies we've contributed to the TCGA effort and been grateful for that opportunity we contributed rather differently to a project or so in Canada listed here and as we've gone through this exercise of profiling these malignancies using various modalities and analyzing the resultant sequences I think the field has learned a great deal about the the anatomy of cancer and first and foremost what has been revealed I think as a consequence of all the sequencing is the incredible heterogeneity of malignancy heterogeneity from a variety of different perspectives clinical outcomes heterogeneity in terms of differences in clinical behavior and metastatic nodes within the same individual and heterogeneity from within a single mass and I just cartooned what I hope will will make my point for me so tumors really are as we've learned communities of cells and so you can think of a community of cells different color different genotype the surrounding micro environment different cell types different functions each color representing a different function or different genotype and this this milieu of cells is really you know cancer and the amazingly high number of cells that one can get an even a relatively small tumor is is illustrative I think of the the fundamental possible complexity of the disease you know you fit so I'm told a hundred million cells in a space about the size of a sugar cube so these communities of cells they talk to each other but they're also dynamic they can co-evolve community of cells under selection maybe by drug certain genotypes may disappear from the population certain genotypes may expand that may be true in the micro environment as well as in the tumor population and so over time this clonal this dynamic clonal reorganization we now know can occur that's important to realize that we actually haven't as a as a community exhaustively sampled this this dynamic clonal evolution most of what we've done has been examining single time points in time and space we've done this this is largely what has been done in the larger public efforts so we have very detailed maps of mutation and commutation from from the space but we do not yet appreciate fully the dynamic nature of disease especially under treatment selection so I think that's an important opportunity that lies before us now and when they were going to try and explore right down to the single cell level using methods now that we have in place to profile single cells at the level of the genome one last point before we close a major emphasis here as you would expect for a cancer agency that services a population of BC BC for those of you that don't know is about twice the the area of California it's only got five million people in it so that's our population and the cancer agency has a mandate for managing that population genomic medicine is very much in vogue here there is a now a stereotypical pipeline which sees whole genome sequencing and RNA seek quite centrally in the pipeline and a system by which these data are reduced to reports on targets and other things of clinical utility there's a discussion about all of these data in the context of multidisciplinary of board and certain examples of these data are being acted on by the oncologist amazingly in our jurisdiction is we have the direct participation now of more than three quarters of all the medical oncologists in the province of British Columbia so there's quite an emphasis here and we look forward to to exploring the space thoroughly in the future this is all enabled by 150 people that participate in our our oncogenomics program the vast majority of which here are medical oncologists with bioinformatics and genome science is being actually a minor minor players in terms of sheer numbers so that's just some thoughts this clonal dynamic nature right down to single cell resolution important for us that's where we think we're headed and and who knows where we'll get to but that'll be an emphasis as well as this of trying to understand in clinical real-time in clinical samples how these these clonal dynamics change with that I'll give you a garbled slide that looked much better when I viewed it on my screen 320 people at the genome center this is the list of them and many funders that have contributed to our cancer program so with that I'll close and apologize once again for the the unusual venue I hope that I could be heard in my breathless account of past history thank you very much for the opportunity to contribute and I guess there may be a few moments for questions at at Jeff's leisure Chris or whoever's running the show at your end thank you thank you very much Mark it was I think it was quite effective even though we're separated by a few miles I invite anyone who has questions to please come to the mic so that Marco can hear us Marco I'll start off with a question so a number of the people you spoke of spent quite large portions of their careers very focused laser focused on developing new methods and yet almost all of them also used those methods among many others to reveal really important insights into biology and often with a very high medical relevance are you have any thoughts on how those ways of thinking in science interact with each other support each other well the technology development angle is absolutely of course critical to everything that we do the the very interesting observation is that you know we quite frequently will buy a piece of hardware to perform a task and then be dramatically disappointed in its inability to execute the thing that we wanted it to do and so in our in our own shop the ability to engineer rethink build these things are absolutely critical and in our experience are in our context anyways best driven from the the end-use and vision I think a good example of this you know that the infrastructure around the fingerprinting most of that was designed by in-house engineers became absolutely critical to our ability to participate in these NHGRI projects another example is the creation of a robot that was capable of running hundreds of RNA samples for purposes of size separating out micro RNAs this allowed us to participate in the TCGA manufacturing I don't know 12 15,000 micro RNA libraries very high-quality libraries with size selected RNA we could not have done that manually so it's incredibly enabling we couldn't we couldn't do without a Jeff any other questions if not thank you once again Marco we very much enjoyed your talk and I hope your recruitments go swimmingly okay thank you all very very much I appreciate the opportunity once again your conference is scheduled to end in two minutes