Harkins: 10ng of PCR-free; at only 15x WGS coverage, get 98% 1x of the genome covered; 84% at >10x (!) #TRICON
6:45pm March 10th 2016 via Hootsuite
Harkins: No need for adapter titration, P7 onto repaired 3' termini, then optional PCR afterward #TRICON
6:43pm March 10th 2016 via Hootsuite
Harkins: Accel-NGS 2S Plus: they actually dephosphorylate first (opposite); repair of both 3' and 5' ends. Then P7/P5 sequential lig #TRICON
Harkins: Reminder: 1 ng = 167 copes, 334 ch copies of any locs. 1% of a het = 3 chromosomal copies #TRICON
6:42pm March 10th 2016 via Hootsuite
Harkins: Conversion rate: type (FFPE, cfDNA, cell lines); fixation; size selection; seq depth #TRICON
6:41pm March 10th 2016 via Hootsuite
Harkins: Using chemagic360 (Perkin Elmer) - 5-20ng from 10mL blood draw. Quality measured by 247/115 ratio of 0.3-0.5 #TRICON
Harkins:cfDNA trace - 170bp. You don't want to see HMW DNA - from the lymphocyte lysis. A two-ALU repeat assay: 115bp and 247bp #TRICON
6:40pm March 10th 2016 via Hootsuite
Harkins: As you get lower inputs the quality lowers as well. Running a service lab: the biggest problem is getting DNA #TRICON
6:38pm March 10th 2016 via Hootsuite
Harkins: He had a front-seat view into sequencing James Watson (on 454). 20ug then. They can now do single-cell methylation #TRICON
Harkins: Even viral and phage metagenomes, very damaged DNA (single-stranded library prep) called Accel-NGS 1S #TRICON
6:36pm March 10th 2016 via Hootsuite
Harkins: Based in Ann Arbor MI, work on all platforms including @pacbio and have a suite of products: WGS, somatic mutations, methyl #TRICON
6:35pm March 10th 2016 via Hootsuite
Tim Harkins (Swift Biosciences) Better library generation, better data for cfDNA NGS applications #TRICON
Q:How about CNV/SNV calling, testing affect on their methods? Drake: Using many methods, not a lot of data yet #TRICON
5:49pm March 10th 2016 via Hootsuite
Drake: Tabulate a feature set, and a classifier. It's a function on the feature data to give a probability of cancer #TRICON
5:32pm March 10th 2016 via Hootsuite
Drake: Applying statistical learning to ctDNA; doing WGS to cfDNA; ability to discriminate from disease or non-disease tissue #TRICON
5:29pm March 10th 2016 via Hootsuite
Adam Drake (Freenome) Deep learning of circulating tumor DNA genome-wide sequencing data yield novel early detection signature #TRICON
5:26pm March 10th 2016 via Hootsuite
Bettegowda: Going to other tissue types: Pap smear material '13 Sci Transl Med Showed data from saliva https://t.co/mvJHKpdqvk
5:22pm March 10th 2016 via Hootsuite
Bettegowda: Post-anti-EGFR treatment, looking at MAPK status in plasma: 70 mutations detected, ave 2.9 muts/tumor #TRICON
5:20pm March 10th 2016 via Hootsuite
Bettegowda: Stage II CRC, looking post-op of ctDNA levels.KRAS codon 12/13 in 206 metastatic cases showed good sens, spec #TRICON
5:19pm March 10th 2016 via Hootsuite
Bettegowda: Describes SafeSeqS, '14 Sci Trans Med https://t.co/LcXa5ykh6U Onto Colorectal ca and CEA marker assoc'd w/survival #TRICON
5:17pm March 10th 2016 via Hootsuite
Chetan Bettegowda (Hopkins) Application of tumor-derived DNA in the management of individuals w/cancer #TRICON
5:15pm March 10th 2016 via Hootsuite
Raghavan: the model of the future - speed of evolution of progress in cancer care #TRICON
3:13pm March 10th 2016 via Hootsuite
Raghavan:Impact - reduced travel and expense for pts, standardization of care improved; accrual to trial has risen 3x; red. Rx cost #TRICON
3:12pm March 10th 2016 via Hootsuite
Raghavan: Consent sheets available across the system; iPhone side-effect reporter app. #TRICON
3:11pm March 10th 2016 via Hootsuite
Raghavan: For molecular testing: gene expression, tumor genetic profiling, protein, blood-based markers, germline. TAPUR/ASCO, Caris #TRICON
3:10pm March 10th 2016 via Hootsuite
Raghavan: Lung ca, then electronic enrollment; look at data on pt and test and physician and related clinical trial, onto a data mgr #TRICON
3:09pm March 10th 2016 via Hootsuite
Raghavan: Info distributed across the system; alerts clinical teams; flowchart shows preferred and acceptable, thus can regulate #TRICON
3:08pm March 10th 2016 via Hootsuite
Raghavan: Created an electronic infrastructure: tumor-spec teams, evidence-based, cost-effective pathways. #TRICON
3:07pm March 10th 2016 via Hootsuite
Raghavan: Built a system where physicians can regulate themselves. Pt entry, registration, EMR; tumor registry; electronic tumor bd #TRICON
3:06pm March 10th 2016 via Hootsuite
Raghavan: building support into their EMR system. Looking at value in healthcare, benefit to pts, clinical benefit, tox, NHB, cost #TRICON
3:05pm March 10th 2016 via Hootsuite
Raghavan: Taken largest hospitals, got them into a tightly regulated network. A single cancer IRB, solving a time problem #TRICON
3:04pm March 10th 2016 via Hootsuite
Raghavan: Their clinical trial participants by minorities - up to 23%! Avoided a pattern of 'skimming', a real-world environment #TRICON
3:03pm March 10th 2016 via Hootsuite
Raghavan: 'Implementing Obamacare in a Red State' - spreading a cancer institute across two states providing care to all #TRICON
3:02pm March 10th 2016 via Hootsuite
Raghavan: Big pattern of costs: poor quality due to uneven distribution, uneven access. #TRICON
Raghavan: Cost of litigation - 'many occult costs' that can't be measured due to indirect effects on medical practice #TRICON
3:01pm March 10th 2016 via Hootsuite
Raghavan: Expensive tech, physician charges, unrealistic community expectations, gov't error, hosp business model, litigation #TRICON
Raghavan: Major problem was inability to get to major centers. 300 miles, 25 facilities. Cost of care: 'exponential increase' #TRICON
3:00pm March 10th 2016 via Hootsuite
Raghavan: 40 hosp, serving about 1M; Levine Cancer Inst est 2011, seeting 15K pts/year; #5 by size in US. A new system built #TRICON
2:58pm March 10th 2016 via Hootsuite
Derek Raghavan (UNC Charlotte) Integrated cancer care - electronic standards and measurementof impact of value vs volume algorithms #TRICON
2:57pm March 10th 2016 via Hootsuite
Schilsky: White House involvement with ASCO at the Moonshot launch #TRICON
2:56pm March 10th 2016 via Hootsuite
Schilsky: 'We think data blocking is a travesty' - on commercial entities locking out any other use of patient EMR data #T'RICON
2:55pm March 10th 2016 via Hootsuite
Schilsky: Challenges - willingness to provide PHI, EHR interoperability; data blocking by commercial enttities; pt education #TRICON
2:54pm March 10th 2016 via Hootsuite
Schilsky: Other plans include comparative effectiveness assessments, patient port for them to access #TRICON
2:53pm March 10th 2016 via Hootsuite
Schilsky: FDA particularly interested in risk stratification, and addressing outcomes following off-label Rx #TRICON
Schilsky: Plans: incorporate genomic data; assess clinical trial eligibility; create long-term treatment plans; risk stratification #TRICON
2:52pm March 10th 2016 via Hootsuite
Schilsky: W/in a given practice, pt is ID'd; outside, de-ID'd. Emphasis now is leading records. 1M by June '16 ASCO mtg #TRICON
Schilsky: CancerLinQ Insights tool: set filters (type of ca, kind of treatment, many parameters) and then drill down #TRICON
2:51pm March 10th 2016 via Hootsuite
Schilsky: Can drill-down into a pt timeline - Dx, surgery, system therapy, radiotherapy, complications. (Amazing.) #TRICON
2:50pm March 10th 2016 via Hootsuite
Schilsky: Shows tiled UI - and % of pts who achieved quality metrics - by ca type, by end of life care, by pain addressed etc. #TRICON
2:49pm March 10th 2016 via Hootsuite
Schilsky: Practices do not need to obtain informed consent, under conditions (listed). '16 J ASCO ref https://t.co/t8VHsOCycX #TRICON
2:48pm March 10th 2016 via Hootsuite
