CW:1,954 sig SNVs; overlap 97 Refseq genes that are well-known, with add'l novel candidates. "32% fall outside RefSeq annotations" #ASHG14
8:09pm October 20th 2014 via Hootsuite
CW: 180 samples, find 900K transcribed cSNVs, put into HWE calculation. P<0.001 and obs heterzygote freq < expected #ASHG14
8:08pm October 20th 2014 via Hootsuite
CW: Ex: MAF of 20%, difference bet imprinted vs. non-imprinted, cp expected vs. observed #ASHG14
8:07pm October 20th 2014 via Hootsuite
CW:Their method - looking at population to circumvent false positives due to technical limitations. Hardy-Weinberg: allele freq #ASHG14
8:06pm October 20th 2014 via Hootsuite
CW: Expression is maternal or paternal allele. Focus on placenta, look for monoallelic expression via imprinting, biased either way #ASHG14
8:05pm October 20th 2014 via Hootsuite
Corey Watson Mt Sinai : An imprinting map of the human placenta based on the application of a pop. genetics approach to RNAseq data #ASHG14
8:04pm October 20th 2014 via Hootsuite
MK: Roadmap datasets available here: http://t.co/lRX4I7Q5T3 #ASHG14
8:00pm October 20th 2014 via Hootsuite
MK: Apply to cancer: rare events in one ca., are high freq with common pathways (regulatory networks). Used to discover new genes #ASHG14
7:59pm October 20th 2014 via Hootsuite
MK:Now Bayesian model linking SNPs to regulatory networks; ex LDL chol. assoc'd locus #ASHG14
7:58pm October 20th 2014 via Hootsuite
MK:Can now show Enh, DNase, Enh+DNase, Super-enh... implications for T1D across marks #ASHG14
7:57pm October 20th 2014 via Hootsuite
MK:On to ID disease-relevant tissues/cell types. Maps of types across GWAS diseases: patterns. HaploReg: http://t.co/ydLXX1ozOx #ASHG14
7:56pm October 20th 2014 via Hootsuite
MK: Now looking at module-based linking of enhancers and target genes. Modules:Modules 'more robust' = a map of enh/promoter linkage #ASHG14
7:54pm October 20th 2014 via Hootsuite
MK: Have started to impute epigenomic data. Now onto activity patterns; 2.3M DNAse elements. Only ~200 activity patterns #ASHG14
7:53pm October 20th 2014 via Hootsuite
MK: Can realte methylation level to chromatin state. Active and repressed regions show differences. Phylogeny shown #ASHG14
7:52pm October 20th 2014 via Hootsuite
MK: Roadmap includes 100+ tiss/cell types. Project homepage: http://t.co/xCYJTUQKlg a dynamic chart - Cell types, color-coded marks #ASHG14
7:51pm October 20th 2014 via Hootsuite
MK:Epigenomics roadmap - enh regions, promoters, transcribed, repressed regions. Very busy slide, many modifications #ASHG14
7:49pm October 20th 2014 via Hootsuite
Manolis Kellis (Broad): Integration of 111 reference human epigenomes helps interpret the molecular basis of complex traits #ASHG14
7:47pm October 20th 2014 via Hootsuite
FG:Q:Cells for ChIP? A:Lymphoblastoid cell lines, resource for TGP #ASHG14
7:44pm October 20th 2014 via Hootsuite
FG:Found hQTLs enriched in GWAS SNPs, OR=1.22 relative to peaks w/o hQTL #ASHG14
7:42pm October 20th 2014 via Hootsuite
FG: Use ChIA-PET interactions, H3K4me1, H3K27ac, H3K4me3 and RNA. 72% of QTL peaks are disrupted by 1 or more SNPs #ASHG14
7:41pm October 20th 2014 via Hootsuite
FG: >50% of all local-distal links involve 2 enhancers; ~7% w/two TSSs. Perh mediated by proximity. One example ~200kb away, H3K4Me3 #ASH
7:40pm October 20th 2014 via Hootsuite
FG: Local to distal distances: >50kb w/in 2MB. ~15% of all local QTLs affect distal sites >50kb away. One can affect several sites #AS
7:39pm October 20th 2014 via Hootsuite
FG: One example for ZNF695 local example of joint local QTLs: 'over 50% of eQTL's are hQTL's' #ASHG14
7:38pm October 20th 2014 via Hootsuite
FG: Find >40K QTL sites, 10-15% of histone peaks are histone QTLs (have genetic var ass'd with them) #ASHG14
7:36pm October 20th 2014 via Hootsuite
FG: 76 YRI individuals; 3 histone marks. Look at local and distal marks, TF motif disruptions #ASHG14
FG: SNPs is 0.1%, CNVs is 0.8% difference. SNPs can affect TF binding, today's study is chromatin (histone) binding differences #ASHG14
7:35pm October 20th 2014 via Hootsuite
First: Fabian Grubert (Stanford): Genetic Basis and Functional Consequences of Chromatin State Variability across Individuals #ASHG14
7:33pm October 20th 2014 via Hootsuite
RT @GholsonLyon: hearing about 3 CFTR mutation homozygotes with absolutely no disease. As part of the Resilience Project. Yep. #ASHG14
3:40pm October 20th 2014 via Hootsuite
SA:Open questions: are GEDDs a phenomena in other trisomies? Can they be reversed? #ASHG14
3:30pm October 20th 2014 via Hootsuite
SA: Mouse partial trisomy mouse Ts65Dn model, Gene Exp Disreg Domains (GEDD), conserved in mouse fibroblasts #ASHG14
3:23pm October 20th 2014 via Hootsuite
SA: Studied one monozygotic twin, where one has T21, other not; global disregulation 2014 Nature publication http://t.co/lGBVcR8noT #ASHG14
3:21pm October 20th 2014 via Hootsuite
RT @girlscientist: .@andrewsu recruited nonscientists via Amazon Mech Turk to manually curate scientific lit for disease relevance. #ASHG14
3:20pm October 20th 2014 via Hootsuite
SA: Goal: ID the transcriptome differences between trisomy 21 and normal cells, hard due to inter-individual transcriptome variation #ASHG14
3:16pm October 20th 2014 via Hootsuite
Stylianos Antonarakis (Univ Geneva) Domains of genome-wide gene expression dysregulation in Down syndrome #ASHG14
3:15pm October 20th 2014 via Hootsuite
RT @erlichya: #ASHG14's first rule: the session in the next room is more interesting.
YH: Their Hi-C bioinformatics pipeline is available here: http://t.co/175UcDu7Fi #ASHG14
3:11pm October 20th 2014 via Hootsuite
YH: 90-99% of enhancer/target genes are on the same chromosome; distance via Hi-C is 123KB. Closest gene is 3.3KB. #ASHG14
3:10pm October 20th 2014 via Hootsuite
YH: Points to this 2011 NAR ref http://t.co/21Wz6ci13b Reviews Hi-C approach: Lieberman-Aiden 2009 Science http://t.co/HkBCFMAXlw #ASHG14
3:05pm October 20th 2014 via Hootsuite
YH: Can use conservation, histone marks or DHS yields >10K enhancers. Where are the target genes? Only 31 validated so far #ASHG14
3:03pm October 20th 2014 via Hootsuite
YH: How do enhancers function: example of preaxxial polydactyly in 2003 has a single-base var in an enhancer of Shh gene, 1MB away #ASHG14
3:02pm October 20th 2014 via Hootsuite
Yih-Chii Hwang (UPenn) Identification and characterization of enhancer and target gene pairs in mammalian genomes #ASHG14
3:00pm October 20th 2014 via Hootsuite
CF:in vitro granular neuron precursor, shRNA knockdown of Zic1 and looking at effect on neural genes ass'd with development #ASHG14
2:54pm October 20th 2014 via Hootsuite
CF: ZIC=zinc finger in cerebellum TF - function to prevent premature differentiation. Act early, severe KO phenotype #ASHG14
2:51pm October 20th 2014 via Hootsuite
CF:>80% dynamic DHS sites are distal from gene promoters. Opening DHS sites enrich for H3K27ac and H3K4me1 #ASHG14
2:50pm October 20th 2014 via Hootsuite
CF:Mapping key time-points on chr11: day 7,14,60: 11K DHS site open or close across development (about 10% of avail. sites) #ASHG14
2:49pm October 20th 2014 via Hootsuite
CF: Look at mouse cerebellum, cerebellar granular neurons are homogeneous, also post-natal differentiation. Doing DNAse-Seq #ASHG14
2:47pm October 20th 2014 via Hootsuite
Chris Frank (Duke): Chromatin accessibility profiling reveals novel neuronal enhancers and regulatory scheme for ZIC TF's #ASHG14
2:45pm October 20th 2014 via Hootsuite
OD:Summarizing - genetic var's affect chromatin binding by first disrupting TF binding #ASHG14
2:43pm October 20th 2014 via Hootsuite
OD:66% of gene assoc are outside promoters. Now DNA var and chromatin marks: distance is bimodal of eQTLs: outside peak 1KB-1MB #ASHG14
2:39pm October 20th 2014 via Hootsuite
OD:Now peaks and genes: 4.5K ass'd genes (~22%), some marks more highly correlated than others; 99% gene-chromatin correl. positive #ASHG14
2:38pm October 20th 2014 via Hootsuite
