Levy: Q: Links to Clinical Trials db? A: Yes, worldwide links avail on the site #AGBT14
11:53am February 13th 2014 via Hootsuite
Levy: Q: Is the API available publicly? A: No, licensing via GenomeOncology for content #AGBT14
11:51am February 13th 2014 via Hootsuite
Levy (now Morrison): Q:how easy to map to GRCh38? A: A problem but we'll 'have to do it' #AGBT14
11:50am February 13th 2014 via Hootsuite
Levy: 'Actionable target summary'; 'Alteration Failed Testing'. Reporting all variants detection #AGBT14
11:46am February 13th 2014 via Hootsuite
Levy: Helped develop 'genomics workbench' tool (GenomeOncology) for merging MiSeq and PGM data; output a summary report #AGBT14
11:45am February 13th 2014 via Hootsuite
Levy: More than test validation: order entry via EHR, with links to exist within EHR. Major effort, many other reporting functions #AGBT14
11:42am February 13th 2014 via Hootsuite
Levy: Goal highest sens and highest PPV, using both MiSeq and @iontorrent PGM with same gene panels #AGBT14
11:41am February 13th 2014 via Hootsuite
Levy: KB management: diagram illustrates 'actionable for tumor type tested', '... for other tumor type' and 'non-actionable' #AGBT14
11:40am February 13th 2014 via Hootsuite
Levy: RPCI test design Mar '13: used NGS, transloc by FISH, CNV by 'digital detection'. PPV, throughput, cost varies #AGBT14
11:39am February 13th 2014 via Hootsuite
Levy: Co-speaker Carl Morrison (Roswell Park Cancer Inst) brought to stage. API developed for daily content updates #AGBT14
11:37am February 13th 2014 via Hootsuite
Levy: Variance evidence levels, and levels of biomarker representation (gene var, exon, fusion, rearr., amplification, protein) #AGBT14
11:35am February 13th 2014 via Hootsuite
Levy: 'Actionable' means: large effect size, and strong evidence. Four effects: sensitivity, resistance, prognostic, diagnostic #AGBT14
Levy: http://t.co/LISd5gofaE first developed in '10 at Vanderbilt with SNaPshot testing. Now 22 Institutions, 19 cancers #AGBT14
11:33am February 13th 2014 via Hootsuite
Levy: Approach to use MyCancerGenome as a kb of clin relevant var's for design and validation. Phase 1: design, pipeline & report #AGBT
11:32am February 13th 2014 via Hootsuite
Levy: Collab with MyCancerGenome and Roswell Park - how to design an actionable panel, how to interpret and QC, and how to report. #AGBT14
11:30am February 13th 2014 via Hootsuite
Up next: Mia Levy (Vanderbilt): Design and validation of an oncology NGS laboratory-developed test focused on actionable mutations #AGBT14
11:28am February 13th 2014 via Hootsuite
Brooks-Wilson: "All interesting variants were present in unaffected individuals - some of whom may go on to develop disease" #AGBT14
11:23am February 13th 2014 via Hootsuite
Brooks-Wilson: But when 4/5 affected individuals: Found 24 likely candidates #AGBT14
11:20am February 13th 2014 via Hootsuite
Brooks-Wilson: Exomes in familial lymphoid cancer: Of 52 potentially damaging, rare variants in a family no single candidate emerges #AGBT14
I'm going to remember that one for a while. Memorable! RT @erlichya: I immensely enjoyed Nolan Kamitaki talk #AGBT14
11:09am February 13th 2014 via Hootsuite
RT @lexnederbragt: #agbt14 inspired blog post: "Defending 'no-tweeting/blogging' requests at conferences" http://t.co/RMeM4ZEbKW
11:08am February 13th 2014 via Hootsuite
Kamitaki: Q: Purer cell types? A: Second sample can change, but only a 10% variance. #AGBT14
10:33am February 13th 2014 via Hootsuite
Kamitaki: Q: Cell type samples? In culture? A: Found in urine (at lower freq). Could be epith. or low-level plasma DNA passed #AGBT14
10:32am February 13th 2014 via Hootsuite
Kamitaki: Q: What about Proteus, other syndromes? A: Will be adding more (less prevalent). Careful phenotyping done #AGBT14
10:31am February 13th 2014 via Hootsuite
Kamitaki: Q: Function of recurring mut sites? A: 3 of them are most common in cancer, other have looked at it #AGBT14
10:29am February 13th 2014 via Hootsuite
Kamitaki: Mutant alleles are usually present at frequencies of <10%. ddPCR ~2h $3/assay #AGBT14
10:28am February 13th 2014 via Hootsuite
Kamitaki: Early mutation could be lethal, later could be CLOVES, later still could be a single-limb syndrome. #AGBT14
10:27am February 13th 2014 via Hootsuite
Kamitaki: Looking at frequency of alleles, mode around 7% in the tissues tested. #AGBT14
10:26am February 13th 2014 via Hootsuite
PSA: If you are interested in digital PCR check out the QuantStudio 3D in the Camaxas 2 room (near the Marriott lobby) @LIFECorporation...
10:25am February 13th 2014 via Hootsuite
Kamitaki: Five PIK3CA muts were tested in the prior dataset. Could have looked at other gain-of-function. #AGBT14
10:24am February 13th 2014 via Hootsuite
Kamitaki: KTS - 22/27 (82%). LM - 18/20 (90%). FAVA 4/8 (50%). #AGBT14
10:23am February 13th 2014 via Hootsuite
Kamitaki: Ended up finding 26/30 CLOVE pts (87%) had PIK3CA mutations. #AGBT14
Kamitaki: Count of droplets, Poisson correction, looked at 30 CLOVES pts, and had seq data in-hand. Found 1/4 pts had PIK3A #AGBT14
10:22am February 13th 2014 via Hootsuite
Kamitaki: Selected digital PCR, allele-specific PCR. Illustrated method of different fluors, emulsions of ~15K droplets. #AGBT14
10:21am February 13th 2014 via Hootsuite
Kamitaki: Want a 'precise digital count of low-freq candidate alleles in affected tissue'. Frequency in biopsy tissues #AGBT14
10:20am February 13th 2014 via Hootsuite
Kamitaki: PIK3 pathway - cell growth, signalling PIK3CA are gain-of-function in many breast, colon tumors Helical and kinase domains #AGBT14
10:19am February 13th 2014 via Hootsuite
Kamitaki: Q: could disparate hypertrophy disorders be related by different points in development? What genes to focus on? #AGBT14
10:18am February 13th 2014 via Hootsuite
Kamitaki: Investigating somatic mosaicism. Could be in early dev, or a smaller scale. Clear cause cancer. All overgrowth disorders #AGBT14
10:17am February 13th 2014 via Hootsuite
Kamitaki: 4 disparate disorders illustrated: CLOVES, Klippel Trenaunay, Lymphatic malformation, FAVA. All present from birth #AGBT14
10:16am February 13th 2014 via Hootsuite
Up next: Nolan Kamitaki (Harvard): Four proliferative disorders arise overwheliming from somatic mutations in PIK3CA gene #AGBT14
10:15am February 13th 2014 via Hootsuite
Mohlke: A (con't): Could be a lot more variants involved in regulation #AGBT14
10:11am February 13th 2014 via Hootsuite
Mohlke: Q: (Chakravarti) is it one SNP or multiple sites? A: GALNT2 has at least of 4 var's, 1 has 80x xcr activity affect. #AGBT14
Mohlke: #IonTorrent 's own Mario Morken in the credits! (From NHGRI in a prior role.) #AGBT14
10:09am February 13th 2014 via Hootsuite
Mohlke:Uses a threshold of at least five reads per allele to det. whether allelic imbalance in effect for a particular loci. #AGBT14
10:08am February 13th 2014 via Hootsuite
Mohlke: Observes these imbalance effects in GALNT2, differential pull-down dep on variants, xcr activity also affected #AGBT14
10:06am February 13th 2014 via Hootsuite
Mohlke: Allelic imbalance in ChIP: allele diff's will affect peak height; in DNAse footprinting a different patter bet. alleles #AGBT14
10:04am February 13th 2014 via Hootsuite
Mohlke: Also for ChIP peak calling - cp BWA to GSNAP, a difference in peak calling at heterozygous sites. #AGBT14
10:02am February 13th 2014 via Hootsuite
Mohlke: Illustrates a C/A SNP plus another error illustrating need for allele-aware aligners. DNAse-Seq data compares BWA to GSNAP #AGBT14
Mohlke: Using higher throughput methods, laying out allelic imbalance measurements. But aligner mapping biases cause problems #AGBT14
10:00am February 13th 2014 via Hootsuite
Mohlke: Switching to HDL: GALNT2 gene, fine mapped 23 var's; cp to open chromatin, promoter and enh regions: many more var's ID'd #AGBT14
9:59am February 13th 2014 via Hootsuite
