Russell Garlick (SeraCare MA) NGS-based solid tumor iagnostics: the next fronter, RNA Fusions #Tricon
6:55pm February 21st 2017 via Hootsuite
Sebra: Did deep WES (116x, 229x, 323x); all 4 germline confirmed; 16 somatic vars in met confirmed '16 ref https://t.co/SvdlksS7NR #Tricon
6:35pm February 21st 2017 via Hootsuite
Sebra: Orthogonal somatic variant detection from bulk tissue: DNA from blood normal; primary FFPE colon (50% pure); met liver (10%) #Tricon
6:33pm February 21st 2017 via Hootsuite
Sebra: But in 3 high-risk pt samples, >100 novel mutations seen of all types. Only 3 samples, but still significantly large number #Tric
6:32pm February 21st 2017 via Hootsuite
Sebra: Whole-gene BRCA 1/2: still find a 'ton of novel information'. 70 targets 3-5kb each; dbSNP is biased toward exons #Tricon
6:30pm February 21st 2017 via Hootsuite
Sebra: Proved functionally in zebrafish (OTX2, mandible of fish model for human). 'Why we should be pushing these technologies' #Tricon
6:29pm February 21st 2017 via Hootsuite
Sebra: Large SVs related to disease: Goldenhar syndrome, craniofacial disorders. Very rare; 1.3Mb chr 14 duplication in proband #Tricon
6:28pm February 21st 2017 via Hootsuite
Sebra: Now doing extreme cases where there are hundreds of repeats for Fragile X. #Tricon
6:27pm February 21st 2017 via Hootsuite
Sebra: Up to 10kb long, very G-C rich. FMR1 for Fragile X: WT 30 repeats vs 120 repeats for affected data shown #Tricon
6:26pm February 21st 2017 via Hootsuite
Sebra: C9orf72 G4C2 repeat discovered in ALS, complex structure https://t.co/6kLkkOypaQ Also want to get to methylation #Tricon
6:25pm February 21st 2017 via Hootsuite
Sebra: Looking now at trinucleotide repeats w/ a PCR-free target enrichment via SMRTbell and capture adapter #Tricon
6:23pm February 21st 2017 via Hootsuite
Sebra: CYP2D6 is hard to phase with heterozygotes. On Sequel: can multiplex 384 samples; also covered GBA, full-length gene #Tricon
6:22pm February 21st 2017 via Hootsuite
Sebra: Each of us have about 1K SVs that are >2500bp in length. These have yet to be characterized #Tricon
6:21pm February 21st 2017 via Hootsuite
Robert Sebra (Mt Sinai SoM NY) Leveraging long read technologies for developing niche assay with diagnostic potential #Tricon
6:18pm February 21st 2017 via Hootsuite
Qiu: Data interoperability on WES '16 ref https://t.co/fTTWUOjoi3 #Tricon
Qiu: Now onto bioinformatics: '13 ref comparing different WES pipelines https://t.co/oQm0WheAgK #Tricon
6:09pm February 21st 2017 via Hootsuite
Qiu: Did tumor purity mixes, from 40% to 100%; showed sensitivity drop from 98% down to 74% the lower % #Tricon
6:08pm February 21st 2017 via Hootsuite
Qiu: Only 326 called by both; lots of false positives in panels w/o matched normals. #Tricon
6:06pm February 21st 2017 via Hootsuite
Qiu: WES requires matched normals; for targeted not req'd. Same samples, tumor burden almost 4x/MB. Same samples! #Tricon
Qiu: Calculation of mutations per MB - shows a input dependency (200ng has 40% higher mutations cp to 50ng input) #Tricon
6:04pm February 21st 2017 via Hootsuite
Qiu: Showed distribution of failures by input, w/dPCR of input: 10ng - 20ng of amplifiable able to get successful Agilent WES #Tricon
6:03pm February 21st 2017 via Hootsuite
Qiu: Theoretically 1 ng = 33o copies; using FF cell ines, WES from 10ng https://t.co/pvZ5g57EKI #Tricon
6:02pm February 21st 2017 via Hootsuite
Qiu: Measured Ct via qPCR for RNA. Showed chart of 'higher DNA input can compensate for low DNA quality'. Used QFI adjustments #Tricon
5:59pm February 21st 2017 via Hootsuite
Qiu: Also wide variation; also some assay-spec differences. RNA peforms comparably, less variability than DNA #Tricon
5:58pm February 21st 2017 via Hootsuite
Qiu: Normalized per mm2 of tissue slices. Wide variation between yields/mm2. Amplifiability: QFI (Asuragen), RNAseP #Tricon
5:57pm February 21st 2017 via Hootsuite
Qiu: 5 labs send same samples from blocks; they extracted RNA/DNA, 5um slides; 2-5 slides sent. Either QIAamp or cobas for DNA #Tricon
5:56pm February 21st 2017 via Hootsuite
Qiu: FFPE suffers from C->T artifacts. Many different kits with pro's and con's. Objective was to ID CRO capability #Tricon
5:55pm February 21st 2017 via Hootsuite
Qiu: DNA input quant and quality. Library prep etc. FFPE challenges: WES has high failures, due to yield and quality. #Tricon
5:53pm February 21st 2017 via Hootsuite
Qiu: Mutational load assoc' wtih response to immunotherapy; gave a lot of references by cancer type. But detection affected... #Tricon
5:52pm February 21st 2017 via Hootsuite
Qiu: Two impt areas- pre-analytical and bioinformatics. Mendelian disorders, NIPT, infectious disease; complex disorders, oncology #Tricon
5:51pm February 21st 2017 via Hootsuite
Ping Qiu (Merck NJ) Implementing NGS as a clinical assay: FFPE pre-analytical optimization in WES commercial lab selection #Tricon
5:50pm February 21st 2017 via Hootsuite
Great work being done @DanaFarber and @VanAllenLab @cityofhopereported at #Tricon https://t.co/IIETLl9goT
5:46pm February 21st 2017 via Hootsuite
Fernandes: Need to be reported in HGVS nomenclature; proficiency testing (CAP) surveys, one for solid tumors #Tricon
5:44pm February 21st 2017 via Hootsuite
Fernandes: Interpretation and reporting results '17 JMD https://t.co/eeuY8saRII Four tiers of reporting; III=VUS; IV=benign #Tricon
5:43pm February 21st 2017 via Hootsuite
Fernandes: Shows sorted graph of list of amplicons by depth; with threshold cutoff for coverage; makes into final report #Tricon
5:41pm February 21st 2017 via Hootsuite
Fernandes: Analytical sensitivity for SNVs and indels: using diluted cell lines, est. LoD of 1.5% #Tricon
5:39pm February 21st 2017 via Hootsuite
Fernandes: For precision (repeatability) 3x's may not be enough. Repro: 3 different days, 3 diff operators #Tricon
5:37pm February 21st 2017 via Hootsuite
Fernandes: Accuracy for gene fusions: select TMPRSS2-ERG; EML4-ALK; MET exon 14 skipped. #Tricon
5:36pm February 21st 2017 via Hootsuite
Fernandes: VAF by percent, run1 vs run2. ERBB2 ampl can be tested by FISH, WES, CNV, need to provide details on algorithm for calls #Tricon
5:32pm February 21st 2017 via Hootsuite
Fernandes: Resistance var's: EGFR T790M; ALK L1196M, G1202R; PDGFRA D842V #Tricon
5:30pm February 21st 2017 via Hootsuite
Fernandes: EGFR in lung ca: two most common are ex19 del (15bp), also exon 21 L858R. About 80% of mutated lung #Tricon
5:29pm February 21st 2017 via Hootsuite
Fernandes: Analytical validation for SNVs - correlate to results from previously validated assay #Tricon
5:28pm February 21st 2017 via Hootsuite
Fernandes: For 100 gene panel - 10 vars per gene? 'Validate what you can, others are 'rolling - as you run and test'. #Tricon
Fernandes: Also each sample type (FFPE, FF, core biopsies etc). SNVs: 20 target areas. Indels: same. CNVs: 10. 3 for Translocations #Tricon
5:27pm February 21st 2017 via Hootsuite
Fernandes: Validation guidelines for NYS: have to be est 'for each type of variant'; SNVs, indels, CNVs, structural variants. #Tricon
5:26pm February 21st 2017 via Hootsuite
Fernandes: Can do tumor geome, transcriptome, exome. Nice figure from '16 review https://t.co/x5Pl4GSwMW #Tricon
5:25pm February 21st 2017 via Hootsuite
Fernandes: Bioinformatics processes have no standardization. Interpretation - same. Points out that Oncology is the fastest growth #Tricon
5:24pm February 21st 2017 via Hootsuite
Fernandes: Review of NGS process: specimen prep; analytical NGS for those having done it for a while, 'it's pretty good') #Tricon
5:23pm February 21st 2017 via Hootsuite
Fernandes: Nov '16, FDA annc a delay finalizing. Jan '17 discussed releasing paper again. #Tricon
5:22pm February 21st 2017 via Hootsuite
Fernandes: During PMI, FDA to create regulatory guidance. '16 June - draft guidance. @SeraCare blog: https://t.co/QfnhielNze #Tricon
5:20pm February 21st 2017 via Hootsuite
