Sathirapongsasuti: 1M samples, 80% opt-in; 1K phenotypes that GWAS can be run with. 300M phenotypic data pts #ASHG15
11:33am October 10th 2015 via Hootsuite
Fah Sathirapongsasuti (23andMe) A practical guide to drug discovery through phenome-wide association studies #ASHG15
11:32am October 10th 2015 via Hootsuite
Taylor: Q: Field-specific? A:Encourages any field, not restricted to biology. #ASHG15
11:31am October 10th 2015 via Hootsuite
Taylor: Currently invite-only, requests via http://t.co/zCeMzohwZf #ASHG15
11:28am October 10th 2015 via Hootsuite
.@markgenome Agreed that combining online with IRL has been great at #ASHG15, also meeting @thatdnaguy and @edyong209 too.
11:28am October 10th 2015 via Hootsuite in reply to markgenome
Taylor: Can keep bookmarks, create user communities, groups to collaborate. Manual curation will be improved via vote +/- #ASHG15
11:26am October 10th 2015 via Hootsuite
Taylor: Can use it to annotate video; their native language. Website: http://t.co/VRD0eotCxd #ASHG15
11:23am October 10th 2015 via Hootsuite
Taylor: Keywords; relationships between articles; adding semantic meaning to improve discoverability. #ASHG15
11:21am October 10th 2015 via Hootsuite
Taylor: Searching is not trivial; Suppl. material? Figures? Terms? Uses example of 'metagenomics' as a term, and its iterations #ASHG15
11:20am October 10th 2015 via Hootsuite
Taylor: PubMed has 3.2K/day in 2014 (1.18M for '14). 1809 is the oldest; newest in 2016. Keyword subm by authors: some MeSH. #ASHG15
11:19am October 10th 2015 via Hootsuite
Taylor: We are all scientific curators - we search, read, make notes. PubMed fun facts: 20.8M English, 322K 'undetermined' language #ASHG15
11:18am October 10th 2015 via Hootsuite
Todd Taylor (RIKKEN): iCLiKVAL: Open-access tool to add value to scientific literature one annotation at a time via crowdsourcing #ASHG15
11:16am October 10th 2015 via Hootsuite
Fujimoto: '13 Lawrence ref data http://t.co/Gdyat6bAkC and TCGA; along with their random 3D permutation test. Shows some overlap #ASHG15
11:06am October 10th 2015 via Hootsuite
Fujimoto: Example, in DIS3 for multiple myeloma: figure from '11 Nature ref http://t.co/jYbSHModNh #ASHG15
11:03am October 10th 2015 via Hootsuite
Fujimoto: Expect number of mutations in drivers and location of mutations expected to not be random. Hotspots via 3D protein struct #ASHG15
11:01am October 10th 2015 via Hootsuite
Akihiro Fujimoto (RIKKEN Japan): Systematic analysis of mut dist in 3D protein structures identifies cancer driver genes #ASHG15
11:00am October 10th 2015 via Hootsuite
Zhao: Mentions STAG2 '11 ref http://t.co/EOclVz2DS2 Others include AHNAK, DDX3X, COL11A1, FAT4, SYNE1 #ASHG15
10:59am October 10th 2015 via Hootsuite
Zhao: In glioblastoma: driver genes on hypermutation of the inactive X-chr in female cancer genomes #ASHG15
10:57am October 10th 2015 via Hootsuite
Zhao: Their model predicted 5 genes with combined genetic, epigenetic alterations. Onto unusual X-chr effect http://t.co/uEUFIWWzRX #ASHG15
10:55am October 10th 2015 via Hootsuite
Zhao: Describes a 'gene gravity' model work '15 PLOS http://t.co/SXm7EciFYP uses TCGA data both somatic and transcription #ASHG15
10:52am October 10th 2015 via Hootsuite
Zhongming Zhao (Vanderbilt) Insights into somatic mutation-driven cancer genome evol: 3,000 cancer genomes across 9 cancer types #ASHG15
10:49am October 10th 2015 via Hootsuite
MacArthur: Next year >100K; moving onto 3,600 whole genomes; sharing ages; user-friendly tools. #ASHG15
10:45am October 10th 2015 via Hootsuite
MacArthur: extrapolate their syn, missense and LoF at 500K - will observe 12% of all possible missense var's. 7.5% of all poss LoF #ASHG15
10:44am October 10th 2015 via Hootsuite
MacArthur: Found 3,230 genes LoF-intolerant. >80% no known LoF phenotype. #ASHG15
10:43am October 10th 2015 via Hootsuite
MacArthur: Empirical ID of genes subject to strong human constraint: DYNCH1H1, UBR4; clear that het LoF is disease phenotype #ASHG15
MacArthur: Onto mutational constraint: ID genes with significant depletion of variation. '14 ref http://t.co/AQdBcYlJcZ #ASHG15
10:42am October 10th 2015 via Hootsuite
MacArthur: CpG saturation detectable only when samples above 10K. Rising line flattening for CpGs #ASHG15
10:41am October 10th 2015 via Hootsuite
MacArthur: Striking effects - ExAc beginning to saturate CpGs. Proportion of all possible CpG syn subst: 87% found in ExAC! #ASHG15
10:40am October 10th 2015 via Hootsuite
MacArthur: 'Everything old is new again': distribution of mut rates: transversions, transitions, and CpG transitions (most freq) #ASHG15
10:38am October 10th 2015 via Hootsuite
MacArthur: 'indels coming soon'. Add'l 35% SNPs soon too. Value for rare disease filtering: 100 individ's; filter 0.1% ESP vs ExAC #ASHG15
MacArthur: Goal: http://t.co/xWEhLl7xGq can see coverage, freq distribution, confidence. Now: Read visualization. 65% of SNPs #ASHG15
10:36am October 10th 2015 via Hootsuite
MacArthur: Largest collection; 10M coding variants, 'one per every six base pairs'. Most rare and novel. >50% singletons. #ASHG15
10:35am October 10th 2015 via Hootsuite
MacArthur: Subset of 60,706 'reference'; not necessarily disease-free, but quality. >50% EU, but more diversity. #ASHG15
MacArthur: From a variety of GWAS cohorts; reprocessed wtih BWA/Picard. Joint calling all w/GATK3 Haplotype caller #ASHG15
10:34am October 10th 2015 via Hootsuite
MacArthur: 'The best time' for statistical geneics - due to HT sequencing. ExAC consortium; last year ExAC data announced, 92K #ASHG15
10:33am October 10th 2015 via Hootsuite
Dan MacArthur (Broad) Combined analysis of over 60K exomes: Genic constraint, mutational recurrence, and impact on interpretation #ASHG15
10:32am October 10th 2015 via Hootsuite
Small:Q:W/4 SNPs in same direction, evidence of positive selection? A:Not from any measure, although interesting question. #ASHG15
10:08am October 10th 2015 via Hootsuite
Small: More large cells (mark of insulin resistence) present in healthy individuals with risk allele #ASHG15
10:06am October 10th 2015 via Hootsuite
Small: Generated KLF14 shRNA knockdowns; decrease in lipogenesis; affect is female-specific to the risk allele (CC). #ASHG15
10:04am October 10th 2015 via Hootsuite
Small: Trait assoc'n: fasting insulin, HOMA-IR, Triglycerides, HDL-C, and hip-size in women. #ASHG15
10:02am October 10th 2015 via Hootsuite
Small: T2D risk alleles decrease LKF14 exp; did site-directed mutagenesis among 5 SNPs; may not be single causative but as grp #ASHG15
9:59am October 10th 2015 via Hootsuite
Small: KLF15 transcrip. repressor, long haplotype limits fine-mapping. Adipose-spec methylation, find a met-QTL #ASHG15
9:58am October 10th 2015 via Hootsuite
Small: #ASGenetics of T2D - maternal-specific, Kong et al 2009; 850 female twins, KLF14 regulates 385 genes in adipose HG15
9:57am October 10th 2015 via Hootsuite
Karin Small (Kings Coll) Adipose- and maternal- spec reg. var's at KLF14 infl. T2D risk by female-spec effect on adipocyte physiol #ASHG15
9:54am October 10th 2015 via Hootsuite
Redin:Q:Repeats at BCA? A:No particular enrichment of different classes, in their global analysis to-date #ASHG15
9:52am October 10th 2015 via Hootsuite
RT @splon: Redin #ASHG15 - Six patients w/ breakpoint in chromosome 5 in 5q14.3 microdeletion region. All between MEF22 & POLR3G.
9:50am October 10th 2015 via Hootsuite
Redin: MEF2C: Hi-C data suggest model w/close proximity between MEF2C with enhancers; how deletions, transloc and inversions happen #ASHG15
9:49am October 10th 2015 via Hootsuite
Redin: Enrichment for ID, ASD. Functional impact put into classes - evidence (42%), recessive, novel 11%, unchar, intergenic. #ASHG15
9:46am October 10th 2015 via Hootsuite
Redin: 78% of do novo BCAs disrupt at least one gene (44%); 2 genes (21%). Large list of dozens of papers; ID, autism, epilepsy #ASHG15
9:42am October 10th 2015 via Hootsuite
Redin: Most NHEJ (blunt-end) (49%) and microhomology (41%) predominate. Hi-C data suggest involved loci are close spatially #ASHG15
9:41am October 10th 2015 via Hootsuite
