Ellis: Characterization of fragmentation: time-dependent, indep. of DNA input amount 'across several orders of magnitude' #AGBT17
9:45am February 16th 2017 via Hootsuite
Ellis: NEB's newest (to launch soon) Ultra II-FS, using enzymatic frag, end-repair and A-tail in single-tube. #AGBT17
9:44am February 16th 2017 via Hootsuite
Ellis: Ultra II was a bit better but still large drop-off. Major problem was library complexity (# of duplicate reads) #AGBT17
9:43am February 16th 2017 via Hootsuite
Ellis: Tested their current workflow, NEBNext Ultra II LC (input range 100ng down to 0.75ng) for WES v5. Metrics suffered <10ng #AGBT17
Ellis: LCM may only have 150 cells. Need high-quality WGS, without WGA due to bias issues for mutational signatures #AGBT17
9:41am February 16th 2017 via Hootsuite
Ellis: Gaps: pg-level library const at-scale is required. Cancer genome group - want genomics from every human tissue via LCM #AGBT17
9:40am February 16th 2017 via Hootsuite
Ellis: Library costs are exceeding sequencing cost; increased demand for PCR-free (req 1000ng). Low input library const problem #AGBT17
Ellis: 20K samples/mo; 75% of these get made into libraries. Their instrumentation include Covaris LE220, Agilent Bravo; LabChip Gx #AGBT17
9:39am February 16th 2017 via Hootsuite
Peter Ellis (Wellcome Trust UK) Next generation DNA library construction for high-throughput genomics #AGBT17
9:38am February 16th 2017 via Hootsuite
RT @richardmleggett: At #agbt17 @pacbio promising Sequel thrupt will increase 2x this year, 2x next year, 8x at end of '18 making 32x total.
8:57am February 16th 2017 via Hootsuite
ICYMI: My own take on the @nanostringtech Hyb & Seq technology presented yesterday at #AGBT17 https://t.co/buNCoKLIVB
8:56am February 16th 2017 via Hootsuite
Marth: Also plans to use single-cell sequencing to better understand clonal expansion, improve their model #AGBT17
9:30pm February 15th 2017 via Hootsuite
Marth: Used multiple sampling from a TNBC post-mortem sample set, to 'call into question the cancer's organ of origin' in mBC #AGBT17
9:29pm February 15th 2017 via Hootsuite
@coregenomics has published his thoughts on @nanostringtech Hyb & Seq here: https://t.co/AOO3e6QBUV #AGBT17
9:24pm February 15th 2017 via Hootsuite
Marth: 2nd line Rx same pt example, a clear resistance mutation with recommendation to change Rx #AGBT17
9:21pm February 15th 2017 via Hootsuite
Marth: Ultimate goal - to be informative for individual pts. No evidence for chemoresistance in example given #AGBT17
9:20pm February 15th 2017 via Hootsuite
Marth: Have models of a consensus subclone model; confirmed with single-cell sequencing, pre and post-Rx. #AGBT17
9:18pm February 15th 2017 via Hootsuite
Marth: Often no unique solutions, multiple subclone structures. '14 Genome Biol ref https://t.co/X0eTDKekpy #AGBT17
9:15pm February 15th 2017 via Hootsuite
Marth: Illus primary tumor, groups of variants that disappear in relapse. Can reconstruct primary and relapse AFs #AGBT17
9:14pm February 15th 2017 via Hootsuite
Marth: Somatic muts are dynamic, forming evolving grouops of cells (subclones) #AGBT17
9:13pm February 15th 2017 via Hootsuite
Gabor Marth (Univ Utah) Tracking the evolution of a patient's tumor across multiple time points or multisite metastases #AGBT17
9:12pm February 15th 2017 via Hootsuite
Syvanen: Shows gain and loss of driver muts in ALL cells at relpase, detectable as minority clones at diagnosis. #AGBT17
9:07pm February 15th 2017 via Hootsuite
Syvanen: ID driver genes, ID'd 5 novel for ALL. ATRX, RPL10, KMT21D. MutSig/Oncodrive computational screens #AGBT17
9:01pm February 15th 2017 via Hootsuite
Syvanen: Dx to relapse, used Haloplex 872 ca genes. '16 ref https://t.co/3hZRx5dmQz Graph of mutational burden #AGBT17
8:54pm February 15th 2017 via Hootsuite
Syvanen: Shows pie chart of distribution of sub-types. 30% of pts lack cytogenetic subtype #AGBT17
8:53pm February 15th 2017 via Hootsuite
Syvanen: Nordic oncologists established biobanks of samples early (>1K DNA, >500 RNA, BM at diagnosis) #AGBT17
8:51pm February 15th 2017 via Hootsuite
Ann-Christine Syvanen Uppsala Univ: Deep-targeted sequencing in pediatric acute lymphoblastic leukemia unveils distinct mut patterns #AGBT17
8:50pm February 15th 2017 via Hootsuite
Nordland: Built classifiers with known subtypes; with unkn samples didn't get the detected they expected, went back to RNA-Seq #AGBT17
8:37pm February 15th 2017 via Hootsuite
Nordland: Concern about overtreatment, given 95% OS rate. Interested in DNA methylation; looked at 756pts in '13 on 450K array #AGBT17
8:35pm February 15th 2017 via Hootsuite
Nordland: 200 cases/year. Transloc's are the core of subtype Dx. Non-random rearr; corr betw. cytogenetics, outcome #AGBT17
8:33pm February 15th 2017 via Hootsuite
Nordland: Ped ALL most common pediatric malignancy; immature B- or T-cells, 2-5y of age onset. Nordic society: OS >90% #AGBT17
8:32pm February 15th 2017 via Hootsuite
Jessica Nordland (Uppsala Univ): Dissection of the heterogeneous fusion gene landscape in pediatric acute lymphoblastic leukemia #AGBT17
8:31pm February 15th 2017 via Hootsuite
Perera: Also combined with metabolyte analysis of urine; set of lncRNA, mRNAs and metabolites enriched in PCa urine. GOT1 impt cmpd #AGBT17
8:29pm February 15th 2017 via Hootsuite
Perera: Collaborating with @Genomics_Guy from Illumina, helped with pan-cancer enrichment for urine RNA #AGBT17
8:25pm February 15th 2017 via Hootsuite
Perera: Gives overview of prostate ca and insufficiency of both PSA tests, prevalence of prostate cancer, and need for biomarkers #AGBT17
8:23pm February 15th 2017 via Hootsuite
Ranjan Perera (Sanford Burnham FL) Systems level invest. of urine enriched RNA and metabolites for early prostate cancer detection #AGBT17
8:22pm February 15th 2017 via Hootsuite
Boland: Sample input was only 1.5ng which they needed; not that cost-effective for Amazon instance; 2.8d of compute/2 samples #AGBT17
8:03pm February 15th 2017 via Hootsuite
Boland: '15 ref about integration sites https://t.co/DcVFtQLBZq Approach to use 10X with hg19 plus HPV as reference. #AGBT17
8:00pm February 15th 2017 via Hootsuite
Boland: Integrated shows clearly interruption of E2 gene. Cancer w/HPV integration, 1 or 2 sites observed. Each site is unique #AGBT17
7:58pm February 15th 2017 via Hootsuite
Boland: Episomal vs integrated HPV: Replication in episomal, integrated loses E2 fn. Shows 10X Genomics data of IGV alignment to HPV #AGBT17
7:57pm February 15th 2017 via Hootsuite
Boland: HPV 150 related viruses; 8kb in size, E6/E7 known oncogenes. 14 types are >98% of HPV-rel cancers worldwide #AGBT17
7:56pm February 15th 2017 via Hootsuite
Joe Boland (NCI, MD) A novel WGS method to comprehensively map all HPV16 integration events across multiple human genomes #AGBT17
7:55pm February 15th 2017 via Hootsuite
#AGBT17 (For a blog post, I wrote up some notes about NanoString's Hyb & Seq at https://t.co/buNCoKLIVB )
7:53pm February 15th 2017 via Hootsuite
Beechem: Able to pull-down DNA and RNA simultaneously, and sequence in parallel, long reads (up to 33K) and phased reads #AGBT17
7:52pm February 15th 2017 via Hootsuite
Beechem:File format output will be FASTA; 2 independent assemblers under development. Hexamer spectrum, and 'short stack' #AGBT17
7:48pm February 15th 2017 via Hootsuite
Beechem: Thus high coverage / high accuracy regions can be determined by the composition of the 'deck' of probes #AGBT17
7:46pm February 15th 2017 via Hootsuite
Beechem: Coverage map - can 'control the deck'. Since it is targeted, you know where the actionable variants are. #AGBT17
7:45pm February 15th 2017 via Hootsuite
Beechem: Their 4096 probes - separate perfect hexamer complements; separate high Tm vs low Tm. #AGBT17
7:44pm February 15th 2017 via Hootsuite
Beechem: Hexamer sequence, designed to interact with target. Reporter domain (74bp) designed to interact with reporters R1/R2/R3 #AGBT17
7:43pm February 15th 2017 via Hootsuite
Beechem: Targeted, clinical-grade sequencing, depends upon 2 color reporter probe, total of 8 colors (pairs of colors) #AGBT17
7:42pm February 15th 2017 via Hootsuite
