Roche: #amp2015 Nice chart of the consequences of inaccurate results https://t.co/mzybU7cv0q
10:34am November 4th 2015 via Hootsuite
Roche: for EGFR, 2014 reported by van Krieken ECC meeting, 61% accurate. 11 different methodologies, instrument, var quality #AMP2015
10:32am November 4th 2015 via Hootsuite
Roche: 'In many ways, Europe does a lot better than we do in the US'. #AMP2015 2011 ref of KRAS results https://t.co/HSe3ldBeVK
10:31am November 4th 2015 via Hootsuite
Roche: Sanger sequencing: great for blood based genetic testing. Pitfalls - shows what trace looks like at 10%, 15%, 20%, 50% tumor #AMP2015
10:28am November 4th 2015 via Hootsuite
Roche: Shows needle core FFPE; H&E stain has 2% tumor content, happens often. Uses Dalmatian dog spots to illustrate. #AMP2015
10:25am November 4th 2015 via Hootsuite
Roche: Shows data of <50 tumor cells w/IHC+ cells. Commentary on FFPE tissue - heterogeneity, % mutant copies; % amplifiable DNA #AMP2015
10:22am November 4th 2015 via Hootsuite
Roche: Ventana's ALK vs FISH compared; automatable, faster turnaround time, less expensive, ongoing FISH neg IHC pos studies #AMP2015
10:21am November 4th 2015 via Hootsuite
Roche workshop: "Advances in NSCLC testing and treatment" Room 16B 2nd IHC test for ALK approved this year #AMP2015
10:20am November 4th 2015 via Hootsuite
Steve Lincoln (In Vitae) - comments it could also be bioinformatic analysis #AMP2015
10:16am November 4th 2015 via Hootsuite
Konigshofer:A:Could be hyb capture designed for wild-type, also a 6-base insertion marker. Issue will go away #AMP2015
9:48am November 4th 2015 via Hootsuite
de Abreu: Q: Illumina assay had low coverage? A: Noticed a 'huge difference'; may be the panel (hyb capture) #AMP2015
9:47am November 4th 2015 via Hootsuite
Garlick: Q:RNA fusion is biosynthetic? A: Cell lines carrying the translocations #AMP2015
9:46am November 4th 2015 via Hootsuite
Konigshofer:Q: What about mutations that dropped out from P0 to P4? A:Not looked at closely, may be on low-end, subtle signal #AMP2015
9:45am November 4th 2015 via Hootsuite
Konigshofer: Could be a widely-available 'true process control'. Compare H&E stain of PDX vs FFPE-treated cells #AMP2015
9:42am November 4th 2015 via Hootsuite
Konigshofer: KRAS G12C driver mut consistent at 50%; other var freq consistent at Passage 4 (P4). PDX may be appropriate #AMP2015
9:41am November 4th 2015 via Hootsuite
Konigshofer: Amplicon-based assays are resistant to mouse sequence; for hybrid-capture, will be more suscept., can be filtered out #AMP2015
Konigshofer: Using hybridization-based assays, most var's are mouse-derived. 19K variants; 6.5K common to all; but tool can remove #AMP2015
9:39am November 4th 2015 via Hootsuite
Konigshofer: Amplicon-based assay (@illumina TruSeq Custom Amplicon) Mouse-derived var's are about ~10% of var's #AMP2015
9:37am November 4th 2015 via Hootsuite
Konigshofer: Plot of ave var freq against individ var freq; most on the diagonal. Instead of 1% outliers (if homogen) got 5% #AMP2015
9:36am November 4th 2015 via Hootsuite
Konigshofer: Prepared FFPE's from each of 5 P4 PDXs. Determined variant's using @jacksonlab Cancer Treatment Profile assay #AMP2015
9:35am November 4th 2015 via Hootsuite
Konigshofer: CNV analysis - lots of chromosome copy number variation. Asked: are drivers stable? P4 useful as homogeneous material? #AMP2015
9:34am November 4th 2015 via Hootsuite
Konigshofer: PDX's maintain key driver mutations, have similar structure. Illustrates patient, P0 and P4 (passage) #AMP2015
9:33am November 4th 2015 via Hootsuite
Konigahofer: #amp2015 illustrates PDX development and passages. https://t.co/EIs0PfwVzi
9:32am November 4th 2015 via Hootsuite
Konigshofer: PDX: a human tumor grown in immunodeficient mice. Thus - expensive, used for drug development, may be more commutable #AMP2015
9:31am November 4th 2015 via Hootsuite
Konigshofer: NY State guidelines - 50 pt speciments with representative distribution; challenge of rare variants w/limited material #AMP2015
9:29am November 4th 2015 via Hootsuite
Next: Yves Konigshofer (SeraCare project scientist) 'Eval of patient-derived xenograft tumors as suitable validation materials' #AMP2015
9:28am November 4th 2015 via Hootsuite
Garlick: Working with a company called Exaptive, can have peer-review for databasing and comaprison #AMP2015
9:27am November 4th 2015 via Hootsuite
Garlick: Mentions QC software in dev - all QC metrics and variables across NGS workflow - visualization, tracking, discovery #AMP2015
de Abreu: Conclude: complexity, flexible, highly stable, diverse applications (could be used for genotyping assays). #AMP2015
9:26am November 4th 2015 via Hootsuite
de Abreu:50% AF in FFPE - saw between 65-67% for BRAF. Very similar variability between FFPE and Seraseq (biosynthetic) #AMP2015
9:25am November 4th 2015 via Hootsuite
de Abreu:50% AF in FFPE - saw between 65-67% for BRAF. Very similar variability between FFPE and Seraseq (biosynthetic) $AMP2015
de Abreu: All 26 by lab and AF: relatively few outside 1 std dev. AKT1 on AF10 across labs - very consistent #AMP2015
9:24am November 4th 2015 via Hootsuite
de Abreu: Average across AF10 and labs: All labs at AF10; but AF20 3/4 were a little low (both Illumina platform/assays) #AMP2015
9:23am November 4th 2015 via Hootsuite
de Abreu: All data uploaded VCF, BAM, FASTQ; VCF looked at first; of the list of var's, 4 indels, 4 SNVs in HP's, 18 SNVs #AMP2015
9:22am November 4th 2015 via Hootsuite
de Abreu: Goals include performance as routine run control; intra and inter-lab perf; compare to key metrics; eval capture software #AMP2015
9:20am November 4th 2015 via Hootsuite
de Abreu: Is presenting poster on Saturday, ST111 on this work; 10% and 20% AF's. 7 sites, different instruments, different assays #AMP2015
9:19am November 4th 2015 via Hootsuite
de Abreu: An ideal control: high complexity; many different mutation types; stable; many different AFs. Describes Seraseq Mut Mix #AMP2015
de Abreu: They can get control material <100pM or sample <100pM; if any <100pM run fails. Onto coverage: >500x in smpl or ctrl #
9:17am November 4th 2015 via Hootsuite
de Abreu: 1 for library, one for emulsification/enrichment, one for data analysis. Drift - in library quant >100pM #AMP2015
9:16am November 4th 2015 via Hootsuite
de Abreu: Also confirmatory tests; QC samples, proficiency test also. Illus. 4 steps wet bench, 2 steps for dry work. 3 QC for DNA #AMP2015
9:15am November 4th 2015 via Hootsuite
de Abreu: ~1500 clinical on solid tumor, ~100 on myeloid. Lab QC, NGS has workflow QC, control material, prof test. #AMP2015
9:14am November 4th 2015 via Hootsuite
Next: Francine de Abreu in Greg Tsongalis' lab at Dartmouth-Hitchcock. CLIA and CAP-certified. Solid Tumor on PGM; Myeloid on MiSeq #AMP2015
9:13am November 4th 2015 via Hootsuite
Garlick: Mentions GetRM meeting yesterday, discussing reference material; SeraSeq Mutation Mix AF20 and AF10, mixed. #AMP2015
9:12am November 4th 2015 via Hootsuite
Garlick: 'without jeopardizing innovation or access to patient care' #AMP2015
9:11am November 4th 2015 via Hootsuite
Garlick: "@seracare has been at AMP for 15 years" Better tools - quotes AMP CLIA Modernization proposal: assurance of quality #AMP2015
9:10am November 4th 2015 via Hootsuite
Garlick: ALK, RET, ROS partners, also NTRK1/3, FGF3' and TFG. Shows read-span data. Both ArcherDx and Ion AmpliSeq RNA confirmed #AMP2015
9:09am November 4th 2015 via Hootsuite
Garlick: RNA fusion prototype: 12-plex full process control (RNAs in FFPE-cells), scalable, fully customizable, any platform #AMP2015
9:08am November 4th 2015 via Hootsuite
Garlick: Lists 5 mutations, any mutation, any multiplex, any frequency, proprietary stabilization. Shows linearity with SwiftBio's #AMP2015
9:07am November 4th 2015 via Hootsuite
Garlick: ctDNA reference, similar to NIPT material: start with cell lines, fragment and blend, stabilize and dPCR formulate. #AMP2015
9:06am November 4th 2015 via Hootsuite
Garlick: Mutation mix at different allele freq's (down to 1%, aneuploidy reference material (T21, T18, T13), also HIV drug resist #AMP2015
9:05am November 4th 2015 via Hootsuite
