Yandell: Looking for statistically dis-similar variants; damaged genes from a only a trio of samples. Manhattan plot of burden #AMP2015
8:44am November 7th 2015 via Hootsuite
Yandell: With WGS/WES, challenges remain. Developed VAAST tool #AMP2015 ref: https://t.co/87T2AUnp1s Sounds like BLAST for a reason...
8:43am November 7th 2015 via Hootsuite
Yandell: Common diseases caused by rare alleles: burden tests a possible solution. Other problem is Bonferroni correction. #AMP2015
8:41am November 7th 2015 via Hootsuite
Yandell: Manhattan plot figure: 2012 Nature Rev Genet https://t.co/F4vpgVgYT8 #AMP2015
8:40am November 7th 2015 via Hootsuite
Yandell: Missing heritability may be many different rare alleles hiding not in LD with marker SNPs. #AMP2015
8:39am November 7th 2015 via Hootsuite
Yandell: Mentions haplotype LD information to pick up a signal; illustrates w/a Manhattan plot #AMP2015
Yandell: 'We are entering the end of the age of the SNP chip and GWAS'. Explains case/control approach. #AMP2015
8:37am November 7th 2015 via Hootsuite
Yandell:Busy urban hospital for pediatric emergency dx, large families for research and family-based medicine 12M records to 1800 #AMP2015
8:35am November 7th 2015 via Hootsuite
Yandell: (Discloses his affiliation w/Omcia). Utah Genome Project is a local, state and federal-supported project #AMP2015
Mark Yandell (Univ Utah) Combining phenotype and genotype for better precision medicine #AMP2015
8:33am November 7th 2015 via Hootsuite
Mungall: Comment from aud: Concern about justification WGS for everyone, and led astray with lower-order organisms #AMP2015
8:32am November 7th 2015 via Hootsuite
.@ariv2013 I like your term 'next generation socializing' over Twitter at #AMP2015 - wish more were active here. Maybe next year...
8:31am November 7th 2015 via Hootsuite in reply to ariv2013
Mungall: Q:a-synuclein in songbird but no fn? A:Yes there are complexities and challenges. #AMP2015
8:29am November 7th 2015 via Hootsuite
Mungall: Acknowledgements include Haendel at OHSU, Smedley at Sanger, Robinson at Charite, Hochheiser at UPitt #AMP2015
8:27am November 7th 2015 via Hootsuite
Mungall: 'A wealth of knowledge in model organisms'. 'We need all the organisms'. Depth / breadth of knowledge is growing #AMP2015
Mungall: 'Phenotypes are crucial for precision medicine'. Variant interpretation need more than genome data. #AMP2015
8:26am November 7th 2015 via Hootsuite
Mungall: Challenge - can text be mined for phenotypes? #AMP2015 Only 40% but a start '15 ref https://t.co/jUVy9czDfX
8:25am November 7th 2015 via Hootsuite
Mungall: Are working on scaling up 'at time of publication'; CSHL Molecular Case studies will have HPO terms, easier for filtering #AMP2015
8:23am November 7th 2015 via Hootsuite
Mungall: From exomes to genomes, bar chart comparing number of associations (Smedley et al in press data). #AMP2015
Mungall: Built tool called exomiser https://t.co/dS7FCW5dKB applying tool to human phenotypes Bone et al Genet Med in press #AMP2015
8:22am November 7th 2015 via Hootsuite
Mungall: All by all comparison, tool called PhenoDigm '13 ref https://t.co/7WL8SYgoHw set of phenotypes by genes and species #AMP2015
8:20am November 7th 2015 via Hootsuite
Mungall: Uberon is their multi-species ontology. '12 Genome Biol ref https://t.co/ncqfnTvObX going cross-spp #AMP2015
8:19am November 7th 2015 via Hootsuite
Mungall: Have specific concepts, arranged as a graph. Published in '13 NAR ref https://t.co/AiU0lePFV4 Human Phenotype Ontology #AMP2015
8:17am November 7th 2015 via Hootsuite
Mungall: Natural language - thick hand skin = ulcerated paws in mice = palmoplantar kyperkeratosis to a physician. (a problem) #AMP2015
8:15am November 7th 2015 via Hootsuite
Mungall: Genotype info has a computable language - g.241T>C for example - techniques developed since early HGP. Phenotypes? #AMP2015
Mungall: Chart by data-source: Panther is top, at 2.7M, ClinVar 312K etc. Phenotypes from 60 metazoan spp. #AMP2015
8:13am November 7th 2015 via Hootsuite
Mungall: Monarch is a one-stop shop, emph on rare and Mendelian diseases, but branching out. Shows ass'ns by data source, 4.1M #AMP2015
8:12am November 7th 2015 via Hootsuite
Mungall: alpha-synuclein (SNCA), shows OMIM, MGI (Mouse), FlyBase data; a lot more depth in function #AMP2015
8:11am November 7th 2015 via Hootsuite
Mungall: But add Mouse, Rat, others and get to 90% of gene coverage (from 40% for human alone). #AMP2015
8:10am November 7th 2015 via Hootsuite
Mungall: Mention of ClinVar, db's of orthologous genes in model organisms. #AMP2015 Monarch Initiative https://t.co/JxNmbFxrhX Human: 40%
8:09am November 7th 2015 via Hootsuite
Mungall: But the phenotype is being lost; but phenome-based interpretation has obstacles: no one source, how to search, match #AMP2015
8:07am November 7th 2015 via Hootsuite
Mungall: Common var's unlikely to be pathogenic: the problem of too many var's, or filtering too strict and var lost; or intergenic #AMP2015
8:05am November 7th 2015 via Hootsuite
Mungall: Mentions @monarchinit and @chrismungall on his opening. Shares a typical WGS/WES, filtering, cp to genomic data, dx, tx #AMP2015
8:04am November 7th 2015 via Hootsuite
Christopher Mungall (Lawrence Berkeley Nat'l Lab CA) Computing on phenotypes across scale and species #AMP2015
8:02am November 7th 2015 via Hootsuite
RT @CompBiolPapers: Design and Analysis of Single-Cell Sequencing Experiments. https://t.co/cTjYRimXlD
7:25am November 7th 2015 via Hootsuite
Google Searches Put Consumers at Risk - The Atlantic https://t.co/Ay6rGd7tLw
6:14am November 7th 2015 via Hootsuite
Daber: Showed uniformity of count; won't cover data analytics for this talk. Moving toward toward validity, an analytical claim #AMP2015
12:40pm November 6th 2015 via Hootsuite
Daber: Lngth of hyb probe and melting temps; showed GC% by coverage. When improved - overbalanced at high GC, adj down #AMP2015
12:39pm November 6th 2015 via Hootsuite
Daber: On the capture side: indep of G-C content, to increase uniformity. Hybrid capture probe spacing (balancing); strand alt #AMP2015
12:38pm November 6th 2015 via Hootsuite
Daber: Increasing PEG, up to 50% increase. Then adding 1mM HCC #AMP2015 1985 NAR ref https://t.co/aXD8TWhPvS using hexamine cobalt chloride
12:37pm November 6th 2015 via Hootsuite
Daber: From 1K molecules, 100 ligated; of which 5 (5%) are abnormal. So ligation efficiency important. #AMP2015
12:36pm November 6th 2015 via Hootsuite
Daber: Their focus - adapter ligation, capture post-PCR, and analysis. Work on library diversity: ligation is typically 10-30% #AMP2015
12:35pm November 6th 2015 via Hootsuite
Daber: But the cost for WES at 1000x is a problem; also some regions still not covered well. #AMP2015
12:34pm November 6th 2015 via Hootsuite
Daber: Depth of coverage - where the claim can be made. Somatic panels are small for hotspots. Germline, only for low volume #AMP2015
12:33pm November 6th 2015 via Hootsuite
Daber: Clinical claim: 3-4% of epilepsy cases (to find); want to validate. Sampling bias; but can leverage it. NY State claim #AMP2015
12:32pm November 6th 2015 via Hootsuite
Daber: 46% of these had alterations in <20% of the nuclei tested. Measuring down to 1-2%. Shows IGV data, showing 1.5% mosaic #AMP2015
12:31pm November 6th 2015 via Hootsuite
Daber: An admixture of different genomic complements; 1.7% of patients so far (122/7,204) via SNP arrays. 16.4% of abnormal finding #AMP2015
12:29pm November 6th 2015 via Hootsuite
Daber: Finding mosaicism: post-zygotic; often the same pathways and genes for dev disorders. Either postzygotic error, or rescue #AMP2015
Daber: Mosaicism 'is a passion of mine'; now in a unique situation for both oncology and germline testing. Somatic genetics #AMP2015
12:28pm November 6th 2015 via Hootsuite
Robert Daber (GeneDx): Hybrid capture probe design to improve coverage uniformity and sens. of mosaic mut. det. in epilepsy cases #AMP2015
12:27pm November 6th 2015 via Hootsuite
