Farnham: Suspected another enhancer nearby - as the tag SNPs clustered together. (Work ongoing) #ncistg14

4:04pm March 20th 2014 via Twitter Web Client

Farnham: And? None of them changed! Effect was 6.5MB to 2MB away! Many cross-chromosome. #ncistg14

4:03pm March 20th 2014 via Twitter Web Client

Farnham: Showed data for a given enhancer, comparing promoter and enhancer marks, overlaying ENCODE ChIP-Seq data! (wow) #ncistg14

4:02pm March 20th 2014 via Twitter Web Client

Farnham: Then RNA-Seq analysis and 3C and 4C analysis to show effect on transcription #ncistg14

4:00pm March 20th 2014 via Twitter Web Client

Farnham: Used both methods to generate enhancer deletions. Verify via ChIP-PCR, can GFP sort, test deletions with ChIP-PCR #ncistg14

4:00pm March 20th 2014 via Twitter Web Client

Farnham: Now TALENs and CRISPRs explained and differentiated. Heterodimer / FokI recognition, vs. guide RNAs #ncistg14

3:59pm March 20th 2014 via Twitter Web Client

Farnham: Enhancers hard to study - either orientation also long-range effects. ENCODE 2012 paper, only 27% affects adjacent gene #ncistg14

3:57pm March 20th 2014 via Twitter Web Client

Farnham: Removed enhancers on the edges of ChIP peaks; determined 7 with one enhancer that a SNP impacted a known enhancer #ncistg14

3:56pm March 20th 2014 via Twitter Web Client

Farnham: Limiting r2>0.5, found 75 'CRC Risk Enhancers' but only bioinformatically-determined. Visually inspected #ncistg14

3:55pm March 20th 2014 via Twitter Web Client

Farnham: Many SNPs in low LD, and others cluster together. One example - four SNPs in the middle of an active enhancer mark #ncistg14

3:54pm March 20th 2014 via Twitter Web Client

Farnham: From 116K to 4.1K to 746 to 370 unique to normal SNPs. Plot of r2 to distance to tag SNP #ncistg14

3:53pm March 20th 2014 via Twitter Web Client

Farnham: Use FunciSNP (R package) http://t.co/7o4asubEIY connect H3K27Ac peak file to correlated SNPs extracted #ncistg14

3:52pm March 20th 2014 via Twitter Web Client

Farnham: CRC cell line is HCT116 (she did that part of ENCODE), colon sigmoid also produced at Scripps (Ecker, Ren) #ncistg14

3:51pm March 20th 2014 via Twitter Web Client

Farnham: Fn element is enhancers, using H3K27Ac active enhancer histone mark. Enhancers are cell-type specific #ncistg14

3:49pm March 20th 2014 via Twitter Web Client

Farnham: Chose 25 SNPs, only 1 synonymous, all others intergenic. Tag SNPs looking for functional element in high LD. #ncistg14

3:49pm March 20th 2014 via Twitter Web Client

Farnham: For CRC (colorectal cancer), look at risk alleles from GWAS, look at risk-ass'd enhancers, and then 'toggle switches' #ncistg14

3:47pm March 20th 2014 via Twitter Web Client

Farnham: Looking at Colon ca risk, looking at genetics, epigenomics, and genome engineering #ncistg14

3:46pm March 20th 2014 via Twitter Web Client

Peggy Farnham, USC: Using GWAS, Epigenomics, and Genome Engineering to Characterize Cancer-Associated Enhancers #ncistg14

3:45pm March 20th 2014 via Twitter Web Client

Prokunina: Still there are new genes to find; GWAS an important tool (still) for discovery #ncistg14

3:42pm March 20th 2014 via Twitter Web Client

Prokunina: Looked at the geographic distribution of TT allele frequency - strong negative selection after new world migration #ncistg14

3:40pm March 20th 2014 via Twitter Web Client

Prokunina: IFNa treatment is $1000/dose (!) requiring 24 weeks. Publication in J Infect Dis http://t.co/XND0oxyzgA #ncistg14

3:39pm March 20th 2014 via Twitter Web Client

Prokunina: deltaG is 60% in Africans, 30% Europeans, 0-5% Asians. Strongest predictor for spont. clearance of HCV #ncistg14

3:37pm March 20th 2014 via Twitter Web Client

Prokunina: IFNL4 is induced via deltaG allele, vs. wt TT allele, only induced when virus infected. Not predicted before #ncistg14

3:35pm March 20th 2014 via Twitter Web Client

Prokunina: Variants for novel protein causes frameshift and causes ~11 isoforms. Called IFNL4, novel type-II IFN Only 29% homol. #ncistg14

3:34pm March 20th 2014 via Twitter Web Client

Prokunina: Saw induction of IFNL3, with a downstream CTCF (separate transcriptional unit), novel protein. Splice isoforms ID'd #ncistg14

3:32pm March 20th 2014 via Twitter Web Client

Prokunina: SNP ass'd with spontaneous HCV clearance in IFNL3 gene. Did RNA-Seq on viral infectious mimic on liver cells #ncistg14

3:30pm March 20th 2014 via Twitter Web Client

Prokunina: Another area of interest is genetics of infection - Hepatitus C (HCV). 1/6 cancers caused by infections #ncistg14

3:26pm March 20th 2014 via Twitter Web Client

Prokunina: Another loci ass'd with bladder ca is Cyclin E. Accel cell cycle, increases instability, incr. alt splicing form #ncistg14

3:25pm March 20th 2014 via Twitter Web Client

Prokunina: Found multiple loci only assoc'd with aggressive disease in CCNE1 region on Chr19 #ncistg14

3:21pm March 20th 2014 via Twitter Web Client

Prokunina: Divided phenotypes into aggressive and non-aggressive forms, did fine mapping of 1.8K and 1.9K cases respectively #ncistg14

3:20pm March 20th 2014 via Twitter Web Client

Prokunina: Baldder ca - high cost burden, survival also high. NCI had two GWAS 3.5k, 2.4k cases, 13 loci ID'd. #ncistg14

3:19pm March 20th 2014 via Twitter Web Client

Prokunina: Looking for germ-line cancer susceptibility. Bladder ca has environmental suscept. (smoking, pollutants) #ncistg14

3:15pm March 20th 2014 via Twitter Web Client

Ludmila Prokunina-Olsson, NCI From Genetics to Translational Genomics and Clinical Implementation of GWAS Findings #ncistg14

3:11pm March 20th 2014 via Twitter Web Client

Gayther: Connection of GWAS to novel therapeutics is clear; also integrate germ line and somatic genetics for cancer transl. #ncistg14

2:43pm March 20th 2014 via Twitter Web Client

Gayther: Germline genetics of risk stratification is impt. GWAS for biomarker discovery and early screening too. #ncistg14

2:42pm March 20th 2014 via Twitter Web Client

Gayther: On the 'translational potential' of GWAS - risk prediction is hard for individuals now, but will be impt in the future #ncistg14

2:41pm March 20th 2014 via Twitter Web Client

Gayther: CRISPER-cas9 used to UCA1 lncRNA to test a deletion. "Incredibly powerful" Readout is transcription of gene #ncistg14

2:39pm March 20th 2014 via Twitter Web Client

Gayther: Genome editing using TALENs to see what deletion effect on phenotype. ColoR Ca risk SNP deleted, affects phenotype #ncistg14

2:38pm March 20th 2014 via Twitter Web Client

Gayther: Using 3C for looping DNA interactions (1:1) or 4C (1:all) or Hi-C (all-to-all). Hi-C is too much for them (handful loci) #ncistg14

2:36pm March 20th 2014 via Twitter Web Client

Gayther: TCGA-eQTL analysis, can knockdown or over express, for functional validation at an ovarian cancer risk loci #ncistg14

2:31pm March 20th 2014 via Twitter Web Client

Gayther: Circos plot of 'Ovarian Cancer Riskome' - see overlaps of SNPs and regulatory elements #ncistg14

2:29pm March 20th 2014 via Twitter Web Client

Gayther: Looking at an ENCODE track for tissue-specific gene regulation at particular loci. "Absolutely critical we do that" #ncistg14

2:27pm March 20th 2014 via Twitter Web Client

Gayther: Fine mapping success story - 19 ovarian cancer loci in 16kb. #ncistg14

2:26pm March 20th 2014 via Twitter Web Client

Gayther: Consortia: Genetic Assoc Mechanisms in Oncology (GAME-ON) http://t.co/fOqwNOk60i #ncistg14

2:24pm March 20th 2014 via Twitter Web Client

Gayther: Fine-mapping is critical; need to look at non-coding regulatory genome; use 3C and eQTLs; genome editing; assays too #ncistg14

2:23pm March 20th 2014 via Twitter Web Client

Gayther: Going from single SNP to determine what the candidate gene or regulatory element is http://t.co/GLo8vzo6yI #ncistg14

2:21pm March 20th 2014 via Twitter Web Client

Simon Gayther, USC Functional Characterization of Ovarian Cancer Susceptibility Loci Identified by GWAS #ncistg14

2:14pm March 20th 2014 via Twitter Web Client

Landi: Tested methylation in cis against top list of GWAS-identified SNPs, 6/7 had strong association. #ncistg14

2:04pm March 20th 2014 via Twitter Web Client

Landi: Also looked at trans - 585 trans-meQTLs, 373 genes. ID a master regulator NPIPL1, chrom 16p11 #ncistg14

1:54pm March 20th 2014 via Twitter Web Client

Landi: 34.3K cis-meQTLs map to 9,330 genes. Adjusted for sex, age, plate, population stratification + methylation PCA #ncistg14

1:52pm March 20th 2014 via Twitter Web Client