Garlick: For LDTs, CLIA says every assay needs a control. Clin labs are under pressure to reduce costs; QC challenges the same. #TCGC15

3:43pm June 23rd 2015 via Hootsuite

Garlick: Today focuses on NGS; new knowledge, new tests. Still needs quality control to assess drift; principles 'don't change' #TCGC15

3:42pm June 23rd 2015 via Hootsuite

Garlick: SeraCare makes products for the IVD industry; started in the '80's to protect the blood supply (HCV, HIV). #TCGC15

3:41pm June 23rd 2015 via Hootsuite

Russell Garlick (@seracare) "Seraseq quality controls for genomic analysis" #TCGC15

3:40pm June 23rd 2015 via Hootsuite

Correll:Q:What if 3 sep reports are ordered? Standardization? A: They see differences in quality across labs, but also similarities #TCGC15

3:26pm June 23rd 2015 via Hootsuite

Correll: Lower barriers to entry to clinical market. He and John Quackenbush founded Genospace (funded by Oracle) #TCGC15

3:10pm June 23rd 2015 via Hootsuite

Correll: Finding more we do understand, also finding more we don't. Clinical labs: use same tech, same genes, similar reports. #TCGC15

3:09pm June 23rd 2015 via Hootsuite

Mick Correll (Genospace) "Clinical laboratory medicine: from bench science to information science" #TCGC15

3:06pm June 23rd 2015 via Hootsuite

Diehn: Summary: ultimately needs a clinical trial to show effectiveness of this technology. #TCGC15

3:05pm June 23rd 2015 via Hootsuite

Diehn: Without prior knowledge of tumor mutation: limit is 0.4% (measured fraction), need to get down to 0.1%. Has paper under rev. #TCGC15

3:02pm June 23rd 2015 via Hootsuite

Diehn: Showed Stage IB pts and follow-up by tracking ctDNA for residual disease, 32 mos 0 pg/mL (detectable at 3 pg/mL) #TCGC15

3:00pm June 23rd 2015 via Hootsuite

Diehn: AF in tumor at 10%; in ctDNA only 1%. Tracked over time. Radiation treatment - scar or tumor? Prog tracked via ctDNA #TCGC15

2:57pm June 23rd 2015 via Hootsuite

Diehn: Relapse pt with T790M not detected by SNaPshot assay (sens at 10%) but found via CAPP-Seq Mut in tumor and ctDNA same #TCGC15

2:55pm June 23rd 2015 via Hootsuite

Diehn: Showed ROC curve, better at stage II-IV lung ca but still good. Monitored ctDNA in metastatic NSCLC Stage IV Nature Med '14 #TCGC15

2:50pm June 23rd 2015 via Hootsuite

Diehn: Can get to 0.01% detection, SNVs, indels, CNV and rearr's. A personalized biomarker. Shows importance of using multi-markers #TCGC15

2:47pm June 23rd 2015 via Hootsuite

Diehn: Methods chart: dPCR, Safe-SeqS, TAm-Seq cp to WGS/WES. They developed CAPP-Seq. ID'd spec regions per class of cancer #TCGC15

2:42pm June 23rd 2015 via Hootsuite

Diehn: ctDNA by tumor-spec muts, can look at methylation, LOH. Challenge is low amt '14 Nature Med http://t.co/xxUF1HP04E #TCGC15

2:39pm June 23rd 2015 via Hootsuite

Diehn: cell-free DNA typically is 80-90% from hematopoietic cells. cfDNA has size distribution mode around 160bp. #TCGC15

2:36pm June 23rd 2015 via Hootsuite

Diehn: Distinguishes between CTCs and ctDNA Fig from '13 ref http://t.co/lEmTtBAEZ0 5ng /mL plasma <2h half-life, changes w/trauma #TCGC1

2:35pm June 23rd 2015 via Hootsuite

Diehn: Liquid biopsies have many potential advantages - ID MRD, prognostic or predictive tests, early detection of recurrence #TCGC15

2:32pm June 23rd 2015 via Hootsuite

Maximilian Diehn (Stanford Univ) "Ultrasensitive detection of circulating tumor DNA by deep sequencing" #TCGC15

2:30pm June 23rd 2015 via Hootsuite

Pasaniuc:Missing heritability: DNAse marks, histone methylation analyzed for enrichment of probability of having causal variants #TCGC15

2:22pm June 23rd 2015 via Hootsuite

Pasaniuc: Looking at ENCODE - overlapping GWAS signals to eQTLs Nicolae 2010 http://t.co/yEsMIeS3Q5 '12 DHS http://t.co/9NVKfz49sI #TCGC15

2:12pm June 23rd 2015 via Hootsuite

.@GenomeNathan TIL 613 is the number of mizvot. I can never know beforehand what I'll learn at a genomics conference! #TCGC15

2:03pm June 23rd 2015 via Hootsuite in reply to GenomeNathan

Pasaniuc: Figure from Goh et al 2012 ref http://t.co/kzD2867ORE 42K iCOGS, 7K BPC3, 10K AAPC studies, all linked to TGP #TCGC15

2:02pm June 23rd 2015 via Hootsuite

Pasaniuc: Pr ca with 100 GWAS loci only explains 33% of familial risk 2014 meta-analysis http://t.co/xlBcoLyb8b #TCGC15

1:59pm June 23rd 2015 via Hootsuite

Bogdan Pasaniuc (UCLA) "Integration of genetic and epigenetic data to understand genetic risk of prostate cancer" #TCGC15

1:57pm June 23rd 2015 via Hootsuite

Jackson: Late-stage lung ca pt, Tarceva resp, relapse, then TreatmentMAP, went from ECOG 3 to 1-2 after rec'd treatmnt #TCGC15

1:22pm June 23rd 2015 via Hootsuite

Jackson: SafetyMAP showed risk of diff treatment arms, report convinced change of treatment. Safety + genomic info to primary pt #TCGC15

1:20pm June 23rd 2015 via Hootsuite

Jackson: Gave ex of woman w/br cancer and Foundation report. SafetyMAP recommended change of treatment, difficult to do in Germany #TCGC15

1:19pm June 23rd 2015 via Hootsuite

Jackson: Open trial info also provided; drug-drug interaction analysis also. Reviewed SafetyMAP info. #TCGC15

1:18pm June 23rd 2015 via Hootsuite

Jackson: Includes reimbursement, safety, approval status information about medication, germline predictors of toxicity #TCGC15

1:14pm June 23rd 2015 via Hootsuite

Jackson: Their system mimics what a physician would do. Report limited to 7pg; focus on top 3 treatment options #TCGC15

1:13pm June 23rd 2015 via Hootsuite

Jackson: Treatment, safety, outcome, insight, risk (all "-MAP") extract info and report out. TreatmentMAP illustrated #TCGC15

1:10pm June 23rd 2015 via Hootsuite

Jackson: Have text data-mining tech est in '08, >23M publications. Uses dictionaries and NLP, curated. Also n=1 studies captured #TCGC15

1:07pm June 23rd 2015 via Hootsuite

Jackson: Molecular Health produced an evidence-driven panel design of 613 genes. Has data warehouse, clin interpretation #TCGC15

1:05pm June 23rd 2015 via Hootsuite

Jackson: For CR ca: EGFR amplification but germline toxicity to Irinotecan. Lastly in lung, EGFR erlotinib but toxic to paclitaxel #TCGC15

1:03pm June 23rd 2015 via Hootsuite

Jackson: Dominant Dx strategy is tumor-spec info on cancer gene seq. Ex of germline mut predicting toxicity to Dabrafenib #TCGC15

1:02pm June 23rd 2015 via Hootsuite

Jackson: Showed drug-drug interaction at different levels - lots of drugs, targets, metabolizing enzymes, then adverse events #TCGC15

1:00pm June 23rd 2015 via Hootsuite

Jackson: Pt extrinsic mechanism: co-medications. Showed distribution of # of drugs/Pt - between 4 and 5 per patient #TCGC15

12:59pm June 23rd 2015 via Hootsuite

Jackson: Tumor intrinsic mechanisms; KRAS and cetuximab in '04 for CR cancer. ID resistance mechanism (many); heterogeneity. #TCGC15

12:56pm June 23rd 2015 via Hootsuite

Jackson: $85B spent on oncology drugs, low likelihood of responding (25% response rate) #TCGC15

12:54pm June 23rd 2015 via Hootsuite

Jackson: Physicians have two people in mind: the patient and Hippocrates. "abstain from whatever is deleterious and mischievous" #TCGC15

12:52pm June 23rd 2015 via Hootsuite

David Jackson (Molecular Health) "Beyond the cancer genome: computational enablement of holistic, evidence-driven pt care" #TCGC15

12:49pm June 23rd 2015 via Hootsuite

Witte:Q:Commercial plans? A:Haven't published this, suspect Kaiser will implement as genotyping inexpensive to do. #TCGC15

12:45pm June 23rd 2015 via Hootsuite

Witte: PSA or adj PSA' and Gleeson score: both highly sig., PSA' had better fit. #TCGC15

12:44pm June 23rd 2015 via Hootsuite

Witte: Showed results of reclass. between controls and cases (PSA >3). Raises question about FP, 4 is no longer cutoff (urologists) #TCGC

12:38pm June 23rd 2015 via Hootsuite

Witte: 7 prev PSA loci, new 24 at p <5x10-8. Summarizes genetic normalization method. Q:Normal? LT followup? A:Yes, but LT in system #TCG

12:36pm June 23rd 2015 via Hootsuite

Witte: Same plot shows previously assoc'd PrCa genes. comparison of PSA to PrCa 'hits': some both, some PrCa only, several PSA only #TCGC15

12:28pm June 23rd 2015 via Hootsuite