RT @AMPath: Burgess: The #FDA has significant work to do before moving to final guidelines. #AMP2015 @michaelcburgess
12:13pm November 7th 2015 via Hootsuite
Nikiforova: Finishes with a few example reports, categories of variant types. #AMP2015
Nikiforova: 30% do not report w/ genes do not meet QC std (i.e. <500x). 70% do put it in w/disclaimer #AMP2015
12:12pm November 7th 2015 via Hootsuite
Nikiforova: Oncologists may have difficulty w/interpretation if AF info is not provided. #AMP2015
12:11pm November 7th 2015 via Hootsuite
Nikiforova: 50% do not report allele frequencies; impt to know b/c it may affect treatment, prognostic, biology of tumor #AMP2015
Nikiforova: Databases: MyCancerGenome almost 80%. 90% do not use commercial template for reporting #AMP2015
12:10pm November 7th 2015 via Hootsuite
Nikiforova: MAF cutoff - majority use 1%. Almost 80% report VUS in their report. 30% likely benign too. #AMP2015
12:09pm November 7th 2015 via Hootsuite
Nikiforova:80% labs do not require normal along w/tumor specimen. What db then used? TGP, dbSNP. Others is NHLBI Exome, custom. #AMP2015
12:05pm November 7th 2015 via Hootsuite
Nikiforova: Literature searches - vast majorit for freq in cancer, functional stuides for activating / LoF. Also: PGx #AMP2015
12:04pm November 7th 2015 via Hootsuite
RT @DrJamesPrescott: Rep Burgess: FDA can't define the problem they are trying to solve. #AMP2015
12:02pm November 7th 2015 via Hootsuite
Nikiforova: Det limit: 85% were 5%. But comments: <1% for cfDNA. Some 'certain hotspots 2%' #AMP2015
12:01pm November 7th 2015 via Hootsuite
.@GholsonLyon Was told by a friend at Omicia there was a surprising amount of interest in their solution (interpreting clinical WES/WGS).
12:00pm November 7th 2015 via Hootsuite in reply to GholsonLyon
Nikiforova: 80% amplicon-based, 35% hyb capture. >200x depth in 93% of labs. 56% report min depth of >500x #AMP2015
11:59am November 7th 2015 via Hootsuite
Nikiforova: 49% heme. Majority run 26-50 genes for solid tumors / heme. lab-developed and commercial both 60% apiece #AMP2015
11:58am November 7th 2015 via Hootsuite
Marina Nikiforova (U Pitt PA): First time report on the survey, will be published. n=66 technical, n=44 general 83% labs ST panels #AMP2015
11:57am November 7th 2015 via Hootsuite
.@GholsonLyon Yes, Yandell spoke this morning (an early-bird session), Mungall is from Lawrence Berkeley, talking about phenome analysis
11:56am November 7th 2015 via Hootsuite in reply to GholsonLyon
Li: AMP organized a working group to look at ISV (interpretation of somatic var's) in somatic conditions, dev list of rec's #AMP2015
11:55am November 7th 2015 via Hootsuite
Marilyn Li (CHOP PA) Guidelines for the interpretation of sequence variants in somatic conditions (cancer) #AMP2015
11:54am November 7th 2015 via Hootsuite
Li: Went back and did targeted at >2000x, found same mutations at 10.6%. #AMP2015 Details '12 Nature Gen https://t.co/u1eeszkE1Q
10:20am November 7th 2015 via Hootsuite
Li: One case, PIK3CA at 4.2% in affected tissue by WES at 167x depth; also in unaffected tissue at 2.7%. But first-pass 5% cutoff! #AMP2015
10:17am November 7th 2015 via Hootsuite
Li: Matches COSMIC mutation frequency, all specific mutations reported in different cancers. Reviews 3 cases #AMP2015
10:15am November 7th 2015 via Hootsuite
Li: 28/50 found mutation; VAF ranged from 1.5% to 49.2%. 25 mutations found in PIK3CA, 14 in PIK3A kinase domain, 6 in helical #AMP2015
10:14am November 7th 2015 via Hootsuite
Li: 50 samples suspected, prenatal to 35yo. 3/4 affected tissue; 1/4 blood. 1 amniotic fluid #AMP2015
10:13am November 7th 2015 via Hootsuite
Li: She shows @iontorrent QC reports, they confirm mutations they report 'with second technology' (explained later) #AMP2015
10:12am November 7th 2015 via Hootsuite
Li: Called 'Overgrowth panel'; uses Ion Reporter; they often use two different SW for var calls. Need 'strict lab QC metrics' #AMP2015
10:11am November 7th 2015 via Hootsuite
Li: Panel is AKT1/2/3, PIK3R2, MTOR and GNAS; entire coding of PIK3CA and PTEN; uses @iontorrent AmpliSeq and PGM; >2000x depth #AMP2015
10:09am November 7th 2015 via Hootsuite
Li: Highlights benefit of targeted, going deeper and seeing lower levels, but of course limitation of breadth of targets #AMP2015
10:08am November 7th 2015 via Hootsuite
Li: Shows Sanger and Pyro traces; CE fragment analysis but 5-10% sensitivity 'not enough' Thus NGS #AMP2015
10:07am November 7th 2015 via Hootsuite
Li: Shows photos of different syndromes; CLOVES is fat tissue. Macrodactyly. 'Very disturbing overgrowth of lower extremities' #AMP2015
10:06am November 7th 2015 via Hootsuite
Li: Lists 12 common overgrowth (Proteus, HIHGHH, MCAP, MPPH, CLOVES, McCune-Albright. Genes AKT1, PIK3CA, MTOR, GNAS #AMP2015
10:04am November 7th 2015 via Hootsuite
Li: Overgrowth can be organ or tissue, focal or diffuse, symmetric or asymmetric; pre- or post-natal. Not inheritable #AMP2015
Li: Tissue mosaicsm; Review '15 Trends Genet https://t.co/IVObURCiWl Can see on one side of body, or different sections #AMP2015
10:03am November 7th 2015 via Hootsuite
Li: Defines - 'karyotypically or genotypically distinct', and derived from single zygote. Could be somatic or gonadal mosaicism #AMP2015
10:01am November 7th 2015 via Hootsuite
Marilyn Li (CHOP PA) Clinical challenges of mosaic overgrowth syndromes using NGS: the power and challenges #AMP2015
10:00am November 7th 2015 via Hootsuite
Biesecker:Q:Usefulness of cfDNA? A:A challenge, may likely work 'for some disorders', much work to be done. Blood, urine, CSF #AMP2015
9:54am November 7th 2015 via Hootsuite
Biesecker: Comment on sampling - homogenizing is a problem, even with skin. Need more work - resolution at single-cell level #AMP2015
Biesecker: Ends w/photo of a 1984 Mardi Gras parade, where the mascot was Proteus the sun-king (title of talk) #AMP2015
9:52am November 7th 2015 via Hootsuite
.@thatdnaguy Good morning Eli! Another Saturday spent live-tweeting...
9:50am November 7th 2015 via Hootsuite in reply to thatdnaguy
Biesecker: Referral for Proteus, negative for AKT1 - but it was from peripheral blood, it was useless. #AMP2015
9:49am November 7th 2015 via Hootsuite
Biesecker: Mosaicism: 'everything is variable'. He compares germline and mosaicism: everything is more interesting / challenging #AMP2015
9:48am November 7th 2015 via Hootsuite
Biesecker: Mosaicism strains concept of a diagnosis: any phenotypic diagnosis w/high PPV doesn't have sensitivity (circular...) #AMP2015
9:47am November 7th 2015 via Hootsuite
Biesecker: Mentions NRC's 'wonderful report' from 2011 (here): https://t.co/QLK97KZF1c Pathophysiology matters. #AMP2015
9:43am November 7th 2015 via Hootsuite
Biesecker: We are going to have to be creative on sampling. Interpretation - sensitivity is critical; neg test not very useful. #AMP2015
9:42am November 7th 2015 via Hootsuite
Biesecker: Uses the term 'modal genotype of the individual" Saliva another, buccal, biopsies, hair bulbs, cfDNA #AMP2015
9:40am November 7th 2015 via Hootsuite
Biesecker: "Mosaicism is the norm." Challenges: sampling - you can't just collect blood. 'Blood is a reasonable proxy" #AMP2015
9:39am November 7th 2015 via Hootsuite
Biesecker: Autosomal dominant, but not in peripheral blood. Mosaicism can be in any organ, any tissue, visible or indivisible. #AMP2015
Biesecker: 2013 ref https://t.co/gb9B8Ps7yJ Cornelia de Lange syndrome - phenotype apparently constitutional but not. #AMP2015
9:38am November 7th 2015 via Hootsuite
Biesecker: And other cases of compensating mutations. Not genetic lethal - low proportion mosaicism, NF1 common for dermatology #AMP2015
9:37am November 7th 2015 via Hootsuite
Biesecker: 'Mosaicism is evolution writ small': there are lethal mutations if constitutional, or viable if constitutional... #AMP2015
9:36am November 7th 2015 via Hootsuite
Biesecker: PIK3CA-related overgrowth syndrome. #AMP2015 Recent reference: https://t.co/NeadCfk1eL
9:34am November 7th 2015 via Hootsuite
