Eichler: Next phase is WGS is where they should put their energy. Go back to Simons Simplex collection #ESHG2016
12:13pm May 22nd 2016 via Hootsuite
Eichler: Transmission of rare LGD within autism families, down to private mutations, with sig odd ratio https://t.co/SbDockweEq #ESHG2016
12:09pm May 22nd 2016 via Hootsuite
Eichler: POGZ ASD/ID subtype - mild ID (70-80) with a happy disposition, tendency toward OCD. https://t.co/sKAjSbFHdv #ESHG2016
12:06pm May 22nd 2016 via Hootsuite
Eichler: These (11 cases ID'd) were not diagnosed before. 89% late onset epilepsy #ESHG2016
12:04pm May 22nd 2016 via Hootsuite
Eichler: Many have been discovered before; many make sense. Now working to re-contact DYRK1A subtype: https://t.co/oSHhMYRNPQ #ESHG2016
12:03pm May 22nd 2016 via Hootsuite
Eichler: With re-contact capability, then follow-up the phenotypes. Found similar maternal bias (54%). 64 sign. ASD/DD genes ID'd #ESHG2016
12:01pm May 22nd 2016 via Hootsuite
Eichler: (Now talking about unpublished data) Approach "Genotype First" '14 Cell ref https://t.co/j7isQGUkCZ #ESHG2016
12:00pm May 22nd 2016 via Hootsuite
Eichler: Using MIPS (Jay Shendure), applied to 500-1000 autism genes. Can prove significance; only $12 for 50 genes in 1 patient #ESHG2016
11:58am May 22nd 2016 via Hootsuite
Eichler: Showed complex gene network diagram from this '14 review https://t.co/WCfoqPxsAn #ESHG2016
11:57am May 22nd 2016 via Hootsuite
Eichler: Genes cluster into pathways and networks; Synapse fn; Wnt signaling; chrom remodeling. #ESHG2016
11:56am May 22nd 2016 via Hootsuite
Eichler: De novo missense - same calculation, 12%. (0.94 proband, 0.82 in sibs) #ESHG2016
11:55am May 22nd 2016 via Hootsuite
Eichler: Est that 42% of de novo truncating mutations contribute to ASD; no compound hets for gene disruption #ESHG2016
Eichler: 2.5K families; 2.1K de novo mut's '14 Nature paper: https://t.co/Ala8oAZlVR By reanalysis: found add'l #ESHG2016
11:53am May 22nd 2016 via Hootsuite
Eichler: Was able to work with 2500 Simons simplex autism quads - w/1 unaffected sibling. Compare types / classes of de novo muts #ESHG2016
11:52am May 22nd 2016 via Hootsuite
Eichler: See a transmission bias - likely transmitted via mothers. But CNVs ID regions but not genes. NGS WES breakthru #ESHG2016
11:51am May 22nd 2016 via Hootsuite
Eichler: Idea - haploinsufficiency; large CNVs account for 14.2% of intell disability; 7-8% for ASD. #ESHG2016
Eichler: Autism - 70-95% identical twins concordant; strong genetic component. Large CNVs - both deletions, dups, de novo & inh #ESHG201
11:50am May 22nd 2016 via Hootsuite
Evan Eichler (Univ Washington US): Using next-generation sequencing to understand the genetics of autism #ESHG2016
11:49am May 22nd 2016 via Hootsuite
Q: Detect FP's? Pajusalu: Used known samples to characterize FPs well. #ESHG2016
10:03am May 22nd 2016 via Hootsuite
Pajusalu: Calling with two tools, Pindel and Platypus. Det 42bp homozygous deletion in PGAP3; 115bp complex indel in another gene #ESHG2016
10:00am May 22nd 2016 via Hootsuite
Sander Pajusalu (Univ Radboud): Pathogenic lon indel ID'd in patients with intellectual disability #ESHG2016
9:59am May 22nd 2016 via Hootsuite
Nissen: Single-exon CNVs are difficult to detect, should be confirmed by another method. Larger datasets for validation needed #ESHG2016
9:57am May 22nd 2016 via Hootsuite
Nissen: CNV det tools" ExomeDepth, Canoes, Codex, Clamms, in-house. Used 52 CNV-pos smples. 41 TP, 11 FN, 1 FP. 78% sens #ESHG2016
9:56am May 22nd 2016 via Hootsuite
Nissen: CNVs hard to validate, they use TruSight Cancer and Agilent Custom (992 genes). #ESHG2016
9:54am May 22nd 2016 via Hootsuite
Third topic: CNV calling in targeted regions. #ESHG2016 Anke Nissen (MGZ, Munich DE) Enrichment varies, introduces noise.
9:52am May 22nd 2016 via Hootsuite
Munroe: WES-first approach chosen; plans to' implement WGS in the future. #ESHG2016
9:42am May 22nd 2016 via Hootsuite
Glen Munroe (Utrecht Netherlands) Traditional Dx - €2296 metabolic cost, €4520 genetic cost. If trio WES: €3600. #ESHG2016
9:41am May 22nd 2016 via Hootsuite
van Nimwegen: Costs concluded to be: targeted €368, WES €989 and WGS €6,157 #ESHG2016
9:39am May 22nd 2016 via Hootsuite
van Nimwegen: Capital costs ~50%, operational ~50%; does not include maintenance. Utilization at 75% WES/WGS, 10% Targeted #ESHG2016
9:36am May 22nd 2016 via Hootsuite
Kirsten van Nimwegen (Radboud Univ Nijmegen) A cost analysis of NGS #ESHG2016
9:35am May 22nd 2016 via Hootsuite
James: 775 cardiac disease genes; showed overlap of genes to their panel; looking at 80 cases and for WGS, 43/80 (59%) #ESHG2016
9:33am May 22nd 2016 via Hootsuite
Paul James (Royal Melbourne Hosp Australia) Inherited cardiac disorders, since '08. Lists somewhat overlapping groups of genes #ESHG2016
9:29am May 22nd 2016 via Hootsuite
Worthey: Concludes WGS is best overall. #ESHG2016
9:27am May 22nd 2016 via Hootsuite
Worthey: Cost: 1.38x the cost of WGS to WES overall. If it is 'reflex', it would req 80% WES MDx success rate to rationalize #ESHG2016
Worthey: Increased MDx success rate: 10-25% higher. Ability to learn more from data over time.#ESHG2016
9:26am May 22nd 2016 via Hootsuite
Worthey: SVs wins for clinical WGS. Also if assays need to change, as assay gets refined. 'This can be a time and resource sink' #ESHG2016
9:25am May 22nd 2016 via Hootsuite
Worthey: Even the best WES do not target all protein coding regions; example, epilepsy genes ~25% poorly covered. #ESHG2016
9:24am May 22nd 2016 via Hootsuite
Liz Worthey (HudsonAlpha AK) Interpretation of WGS/WES 'no more difficult' https://t.co/xWWZ9tXPCx WES poorer in exonic #ESHG2016
9:22am May 22nd 2016 via Hootsuite
Nigro: (He highlights a region of the genome, ARID18, of interest to him but problematic) #ESHG2016
9:20am May 22nd 2016 via Hootsuite
Nigro: WGS may be the better exome, but cost effectiveness another matter. Shows nice comparison of '11 25x WGS to '14 160x WES #ESHG2016
9:19am May 22nd 2016 via Hootsuite
Large panels, WES or WGS? Vincenzo Nigro: Presents chart: Cost, bias, lab work, bioinformatic overhead, CNV calling #ESHG2016
9:18am May 22nd 2016 via Hootsuite
Q2 (S Lincoln): GIAB is only high-confidence calls, only 77%, about 1/4 of the disease genes Biesecker: "Good point." #ESHG2016
9:17am May 22nd 2016 via Hootsuite
Q: About standards, how should labs measure FP rates? Biesecker: Mentions Genome-in-a-bottle and other efforts. #ESHG2016
9:16am May 22nd 2016 via Hootsuite
#ESHG2016 Plenary 'NGS in the Clinic' is several v. short talks highlighting topics. First: Sanger confirmation: time to stop?
9:12am May 22nd 2016 via Hootsuite
Clark: In review, amplification-free method uses 1-5 ug input. Can get enrichment on the order of 38K #ESHG2016
8:18am May 22nd 2016 via Hootsuite
Clark: In same FMR1 sample: CGG repeat heavily methylated. #ESHG2016
Clark: Direct detection of methylation via PacBio: real-time kinetics of polymerase pauses at modified bases. #ESHG2016
8:17am May 22nd 2016 via Hootsuite
Clark: Found a 110bp duplication in FMR1, in addition to 42 CGG repeats (where PCR-only doesn't work) #ESHG2016
8:14am May 22nd 2016 via Hootsuite
Clark: Fragile X; '12 ref https://t.co/K13aar0ckZ exact sequence needed in case of interruptions in regions. Showed >700 repeats #ESHG201
8:13am May 22nd 2016 via Hootsuite
Clark: Use CRISPR-Cas9 after SMRTbell library ligation. gRNA targets region, and re-ligate another SMRTbell. Huntington's repeats #ESHG2016
8:09am May 22nd 2016 via Hootsuite
