Mungall: Several new gene fusions ID'd, '12 Blood https://t.co/2ANMmji0QZ FFPE resource now available. Dev workflow, discovery #AGBT16
8:55pm February 12th 2016 via Hootsuite
Mungall: Survival of MYC or BCL2 and OS, K-M plot shows double-hit disease with very poor survival #AGBT16
8:53pm February 12th 2016 via Hootsuite
Mungall:B-cell lymphomas have recurrent translocations, all involve chr 14 where heavy-chain Ig located #AGBT16
8:52pm February 12th 2016 via Hootsuite
Mungall: B-cell lymphoma, dates back 50y Fig from this '15 Nat Rev Imm https://t.co/Q5IFYKxy2u #AGBT16
8:51pm February 12th 2016 via Hootsuite
Andrew Mungall (BCCA) Detection of genomic rearrangements in archival lymphoma tissues using targeted capture sequencing #AGBT16
8:49pm February 12th 2016 via Hootsuite
Aburatani: oxoG artifact flag (to suppress FFPE-induced artifacts) 15-22% samples affected #AGBT16
8:41pm February 12th 2016 via Hootsuite
Aburatani: DNA damage types can have unique signatures; '14 Nature Rev Gen ref https://t.co/VtIe8GMCdX #AGBT16
8:36pm February 12th 2016 via Hootsuite
.@h2so4hurts Not so surprising - very $$$ to stream video (or even just record it). Believe me, one event was $50K (won't say which one!)
8:34pm February 12th 2016 via Hootsuite in reply to h2so4hurts
Aburatani:HGSOC TP53 is most frequent per TCGA; but PARP inhibitors play a role. 81 cases, WES, CN analysis, also methylation #AGBT16
8:30pm February 12th 2016 via Hootsuite
Hiroyuki Aburatani (Univ Tokyo) Genomic assessment of chemotherapy sensitivity in High Grade Serous Ovarian Cancer #AGBT16
8:28pm February 12th 2016 via Hootsuite
Otto:Q:95% indels, how large? A:From 1 to 40 bases, what's available in cell-lines. Hard to find cell lines with larger #AGBT16
8:25pm February 12th 2016 via Hootsuite
Otto:Q:How do you know unique? A:Barcoding, and unique ends (via shotgun) #AGBT16
8:23pm February 12th 2016 via Hootsuite
Otto: Acknowl. Travis Clark on the lab side, and Mark Kennedy on the computational biology side #AGBT16
Otto: Standard of care biopsies, you need to start with that for the best patient option, but ctDNA 'is a powerful option' #AGBT16
8:22pm February 12th 2016 via Hootsuite
OttoStreck shows ctDNA peak, and KRAS signal (as expected). Tissue vs ctDNA concordance lines up nicely (by disease stage) AGBT16
8:21pm February 12th 2016 via Hootsuite
Otto: Prolonged storage in EDTA, incomplete removal of buffy coat. Bioanalyzer shows no 170bp peak #AGBT16
Otto: Ave plasma 4.5mL, of 168, 69% yielded >2,500x even though low quality plasma samples #AGBT16
8:20pm February 12th 2016 via Hootsuite
Otto: Prelim clinical results: sample procurement of 172, majority had very poor quality plasma. Collection was poor. #AGBT16
8:19pm February 12th 2016 via Hootsuite
Otto: 99.9% PPV for subst, 98% or above for indel, fusions, CNA. Fully autom workflow. Inter- and intra-run concordance >95% #AGBT16
8:18pm February 12th 2016 via Hootsuite
Otto: 100% concordance in orthogonal validation via ddPCR. 87/87 somatic alterations, 47 of these at <5% MAF #AGBT16
8:17pm February 12th 2016 via Hootsuite
Otto: CNA: 39 TP, 3 FP, 1 NN, 93% sens, 98% PPV #AGBT16
8:16pm February 12th 2016 via Hootsuite
Otto: SNV: showed 168 TP, 2 FP, 0 FN, 100% sens, 98.8% PPV #AGBT16
Otto: Demonstrated contamination control and detection. Intra- and inter-run concordance. #AGBT16
8:15pm February 12th 2016 via Hootsuite
Otto: Best practices of FoundationOne - accuracy, PPV for subst, indels, fusions, CNA. Workflow, compatibility and contamination ctl #AGBT16
Otto: Maintain uniformity - normalized coverage plot look very good, down-sampling to 5,000x still 99% covered >2500x #AGBT16
8:14pm February 12th 2016 via Hootsuite
Otto: 90% of samples with >25ng of ctDNA yield >5000x unique exon coverage. #AGBT16
8:13pm February 12th 2016 via Hootsuite
Otto: Library construction and bioinformatics were the major effort. Initial data - overall 50-70% conversion from ctDNA to sequence #AGBT16
Otto: Need 5,000x unique (non-duplicate) coverage. Automating workflow - 18 mos of work. Will share detail 'in near future' #AGBT16
8:12pm February 12th 2016 via Hootsuite
Otto: Pef spec - as many pts as possilbe. PPVs >99% for subst. or >95% for rest (indels, rearr, CNA, CN>8, 10mL plasma 25-200 ng DN
Otto: Other - pts who have relapsed on targeted therapies. #AGBT16
8:11pm February 12th 2016 via Hootsuite
Otto: No way to assay tumor content. 2 key indications: some is better than none. (Biopsy unacceptable risk or unaccessable) #AGBT16
8:10pm February 12th 2016 via Hootsuite
Otto: Formalin fixation - can assay for tumor content. Liquid biopsies - exciting, but indirect interrogation. Apoptosis or necrosis #AGBT16
Otto: 5. Operational - it has to actually work, in a reasonable timeframe. For them - 14d #AGBT16
8:09pm February 12th 2016 via Hootsuite
Otto: 4. Analytic validation - validation results have to be made transparent tot he field. #AGBT16
Otto: Performance metrics, and 'process-matched controls'. 3. Accurate at low MAF, and high spec (PPV, no FPs) #AGBT16
8:08pm February 12th 2016 via Hootsuite
Otto: Five critical features: 1. Comprehensive. "Only one shot" - maximize the Dx yield. 2. High quality - it has to work. #AGBT16
Geoff Otto (Foundation Med): Assessment of the relative clinical utility of ctDNA for detection of actionable genomic alterations #AGBT16
8:07pm February 12th 2016 via Hootsuite
Garvin: 3 homeobox genes implicated. Assoc w/DEG - two genes PTPRN2 and SLC12A8 as measured by PFS. #AGBT16
8:00pm February 12th 2016 via Hootsuite
Garvin: Found a long list in pancreatic ductal adenocarcinoma (PDA). Assoc w/genes impt in PDA - few. But affect pathways #AGBT16
7:58pm February 12th 2016 via Hootsuite
Garvin:1/3 had expression data; GECCO is their pipeline for analysis. Many Q's to answer - CRRs with recurrent non-coding mutations #AGBT16
7:57pm February 12th 2016 via Hootsuite
Garvin: And instead of 'binding site' is 'cis-regulatory regions' (CRR). Via ICGC, 308 pts with WGS, clincial #AGBT16
7:55pm February 12th 2016 via Hootsuite
Garvin: ENCODE has 121 'xcr factors'; but they often do not bind DNA themselves. Should be called 'regulatory proteins' (RPs) #AGBT16
Garvin: .. and chr remodelers. Looking for non-coding muts - promoters, enhancers, introns. #AGBT16
7:54pm February 12th 2016 via Hootsuite
Garvin: Low 5y OS rate; most common form of panc ca. Reg genome: signaling (kinases) xcription factors (repressor/activator) #AGBT16
7:53pm February 12th 2016 via Hootsuite
Tyler Garvin (CSHL): Recurrent noncoding regulatory mutations in pancreatic ductal adenocarcinoma #AGBT16
7:52pm February 12th 2016 via Hootsuite
Seshagiri: BAP1 - BRCA-Assoc Protein 1, epigenetic regulator. OS of TP53 burden - high mutation burden fare worse #AGBT16
7:38pm February 12th 2016 via Hootsuite
Seshagiri: List of significantly mutated mesothelioma - 3 known, 7 novel. NF2: tumor supp, LOF #AGBT16
7:37pm February 12th 2016 via Hootsuite
Seshagiri:675 cell lines of RNA-seq data made available recently '15 Nature Biotech https://t.co/haivnuTAUn #AGBT16
7:36pm February 12th 2016 via Hootsuite
Somasekar Seshagiri (Genentech) Comprehensive analysis of tumors at whole tissue to single cell level #AGBT16
7:34pm February 12th 2016 via Hootsuite
Just having dessert and @nanopore sequencing at #AGBT16 - @pathogenomenick and @mattloose. Good times! https://t.co/kW9y6w7kog
6:48pm February 12th 2016 via Hootsuite
