Lander: 'There is no distinction anymore between a wet and dry biologist' Cannot do computational biology without the wet-lab #NIH_PiDay
12:09pm March 13th 2015 via Hootsuite
Lander: David Baltimore photo in '82, tearing out a room at the Whitehead to devote to computers. #NIH_PiDay
Lander: Concludes with the quantitative nature of biological research; big data for individual variation to individ. cells #NIH_PiDay
12:08pm March 13th 2015 via Hootsuite
Lander: Can predict new patterns of many more genes mathematically, and validate experimentally for developmental biology #NIH_PiDay
Lander:Reconstructing spatial location: 851 cells from dissociated embryo; divide embryo into bins and map 37 landmarks #NIH_PiDay
12:07pm March 13th 2015 via Hootsuite
Lander: Cells transit in-between types. Smoothies vs. fruit salad: "It's nuts to do genomics on smoothies" @fluidigm #NIH_PiDay
12:06pm March 13th 2015 via Hootsuite
Lander:Taking glioblastoma (GBM): 4 types of cells, each all present in different ratios, all discovered on single-cell basis #NIH_PiDay
Lander: (Evan Macosko, Harvard: DropSeq: A Droplet-Based Technology for Single-Cell mRNA-Seq Analysis on a Massive Scale #AGBT15) #NIH_PiDay
12:05pm March 13th 2015 via Hootsuite
Lander: In May '12 18 cells. In July '14 100K cells. Shows microfluidics of DropSeq (presented at #AGBT15) #NIH_PiDay
12:04pm March 13th 2015 via Hootsuite
Lander: 'Single cell genomics has come an incredible way only in the past 18 months' #NIH_PiDay
12:03pm March 13th 2015 via Hootsuite
Lander: Last topic - a cell atlas. Every cell type, every cell state, how they differ by location, transition, and lineage #NIH_PiDay
12:02pm March 13th 2015 via Hootsuite
Lander: Eventually - the healthcare system becomes a learning system. Global Alliance @GA4GH http://t.co/lJFd7J78BE #NIH_PiDay
Lander:1M person cohort - still too small. We need >25K-50K per condition. 'In the end it must be much broader' #NIH_PiDay
12:00pm March 13th 2015 via Hootsuite
Lander: "We need to turn healthcare into a learning system". Medical progress - POTUS support. Website: http://t.co/TZH2E1Fzex #NIH_PiDay
11:59am March 13th 2015 via Hootsuite
Lander:Challenge of Precision Medicine: 'We've only scratched the surface' of common genetic disease. For cancer predisposition.. #NIH_PiDay
11:58am March 13th 2015 via Hootsuite
Lander:There are also nucleotide-specific background mutation rate. 'You can't practice this w/o math'. #NIH_PiDay
11:57am March 13th 2015 via Hootsuite
Lander: Gad Getz' work on cancer. 140 lung ca: too many genes - 'thank God he didn't publish that'. Background mut rate uneven #NIH_PiDay
11:56am March 13th 2015 via Hootsuite
Lander:Another: high-dimensional PCA, for accounting for population structure. Many other advances from math mentioned. #NIH_PiDay
11:55am March 13th 2015 via Hootsuite
Lander:Shows progression at greater and greater samples, to 110K samples. 'The math gave them the faith to keep going' #NIH_PiDay
11:54am March 13th 2015 via Hootsuite
Lander:But math: should have random distribution of significant peaks, which data did not show. 'Should expand the sample'. #NIH_PiDay
11:53am March 13th 2015 via Hootsuite
Lander: Onto GWAS, Macular degeneration (first finding) only from 150 people. For schiz: 6K samples & manhattan plot: nothing #NIH_PiDay
11:52am March 13th 2015 via Hootsuite
Lander:Fractals: Hilbert curves, Minkowski dimensions. 35nm fibers don't exist: at 50kb prevents folding. #NIH_PiDay
11:51am March 13th 2015 via Hootsuite
Lander: (Mentions a model that hasn't been published yet.) The physics of the folding of the genome - can be solved by math. #NIH_PiDay
11:49am March 13th 2015 via Hootsuite
Lander: And CTCF sites are not symmetric; 96% of loops point 'in' to the loop (instead of 25% by chance). #NIH_PiDay
11:47am March 13th 2015 via Hootsuite
Lander:Also conserved across imprinted regions; also correlated to activation of genes; vast maj anchored to CTCF sites #NIH_PiDay
Lander: At 1kb increased contact-map resolution, found only 10K loops in the genome. Conserved across celltypes, across spp. #NIH_PiDay
11:46am March 13th 2015 via Hootsuite
Lander: Instead of an 'equilibrium globule', it's a 'fractal globule'. (NB: Lieberman likens it to a ramen noodle pkg unspooling) #NIH_PiDay
11:44am March 13th 2015 via Hootsuite
Lander:Blue/red regions map to open/closed chromatin. Probability of position in space as a fn of distance. #NIH_PiDay
11:42am March 13th 2015 via Hootsuite
Lander: 3C (chrom conformation capture) in '02 Dekker, Hi-C (Aiden-Lieberman). Nearer/further: 2d chart 'it's Scottish' plaid #NIH_PiDay
11:41am March 13th 2015 via Hootsuite
Lander: Highlights Erez Lieberman's 4D chromosome conformation work, highlighted here in @SciAm: http://t.co/n5ylnkFLfI #NIH_PiDay
11:39am March 13th 2015 via Hootsuite
Lander: Cp to chance: Chen-Stein theory 'under the hood' how we find out about the human genome #NIH_PiDay
11:38am March 13th 2015 via Hootsuite
Lander: 'Thank goodness for dynamic programming algorithms' - allowing to determine blocks of similarity across organisms #NIH_PiDay
11:37am March 13th 2015 via Hootsuite
Lander:Most evolutionary innovation concerns non-coding sequence; innovation often distributed by transposons. >250K instances #NIH_PiDay
11:36am March 13th 2015 via Hootsuite
Lander: In '07 number of genes reduced to 21K. 'Most functional information is non-coding!' ~6% evol conserved; ~1.2% coding #NIH_PiDay
11:34am March 13th 2015 via Hootsuite
Lander:In '01, 35K-120K genes, handful of 'non-coding genes'. 'This was all killed by data'. #NIH_PiDay
11:32am March 13th 2015 via Hootsuite
Lander:Slide of the Mouse genome, then lots of vertebrates. Showed data from '01 "what was true then - I put it on a test" (!) #NIH_PiDay
11:31am March 13th 2015 via Hootsuite
RT @NIH_BD2K: You can’t possibly do data science if the data are behind walls. Said @eric_lander @ #NIH_PiDay http://t.co/yoJR4cZzZ4
11:30am March 13th 2015 via Hootsuite
Lander: Without a cadre of people who could work through the algorithms, the HGP 'would have crashed and burned'. #NIH_PiDay
Lander: Hashing algorithms - no n^2 analyses. Can do lookup tables pre-computed. Graph theory: tracing overlaps. #NIH_PiDay
11:29am March 13th 2015 via Hootsuite
Lander: HGP 'Finished' is a 'technical term' -90% in '01 published; 99.3% '04 published. Story about Science/Nature journals #NIH_PiDay
11:28am March 13th 2015 via Hootsuite
Lander: The sharing of data 'was a real battle; a seminal event in biology' #NIH_PiDay
11:27am March 13th 2015 via Hootsuite
Lander: Onto HGP ('90-'03) - a plan for genetic map (polymorphic landmarks), physical map (clones), sequence, then gene list. #NIH_PiDay
11:26am March 13th 2015 via Hootsuite
RT @NIHPainResearch: #NIH_PiDay Dr. Eric Lander speaking; watch live now! http://t.co/gy060hgAt6 http://t.co/jzUo5agdZH
11:24am March 13th 2015 via Hootsuite
Lander:This stochastic process - a particle on a spring under Brownian motion very similar to setting a LOD score threshold. #NIH_PiDay
Lander:In the 80's 'We knew the information was there, if we could just get at it.' Hard problems: Orenstein-Uhlenbeck diffusion #NIH_PiDay
11:23am March 13th 2015 via Hootsuite
RT @ctorgan: If you love science, immediately go to http://t.co/CObIwPIkQL to watch live videocast of #NIH_PiDay lecture by @eric_lander
11:22am March 13th 2015 via Hootsuite
Lander: 'And a p-value is all the math a biologist knows' All about looking for signals in families, homozyg mapping #NIH_PiDay
11:21am March 13th 2015 via Hootsuite
Lander: Refers to 1980 Botstein paper http://t.co/wzLfPG0cbE for construction of a linkage map. Onto LOD scores #NIH_PiDay
11:20am March 13th 2015 via Hootsuite
RT @ppuigbo: RT @KellyLawhead: With 26 #STEM students at #NIH_PiDay http://t.co/V8X8RfWMqP
11:17am March 13th 2015 via Hootsuite
RT @lisafederer: If you have questions for Dr. Lander please tag them with #NIH_PiDay. we'll get to as many as we can!
11:16am March 13th 2015 via Hootsuite
