Beechem: Concludes with a description of their newest system, the Digital Spatial Profiler (in Tech-Access) #AGBT17
1:39pm February 15th 2017 via Hootsuite
Showing live sequencing at #agbt17 of Hyb & Seq. New blog post: https://t.co/buNCoKLIVB #AGBT17 https://t.co/Bj9iCbf5wK
1:36pm February 15th 2017 via Hootsuite
Parker: #AGBT17 (For details on PREVAIL this from Medivation https://t.co/0TT1bt675A ) PAM 50 captures heterogeneity of BC gene exp
1:29pm February 15th 2017 via Hootsuite
Parker: Looking at basal vs luminal activity in TNBC - PFS doubling that are PREDICT AR+, 5x to recurrance (9w to 40w) #AGBT17
1:26pm February 15th 2017 via Hootsuite
Parker: Working with NanoString's PAM50 classifier, looking at Enzalutamide w/TNBC patients. Did some Monte Carlo and other modeling #AGBT17
1:24pm February 15th 2017 via Hootsuite
New Post: NanoString’s Hyb & Seq single-molecule sequencing platform at #AGBT17 - Next Generation Technologist https://t.co/buNCoKLIVB
1:22pm February 15th 2017 via Hootsuite
Parker: Working with Medivation, Enzalutamide is an inhibitor of AR signaling. #AGBT17
1:16pm February 15th 2017 via Hootsuite
Joel Parker (Univ NC CH) NGS for biomarker discovery and companion Dx development #AGBT17
1:15pm February 15th 2017 via Hootsuite
@coregenomics loading the @nanostringtech instrument at #agbt17 https://t.co/wPV3NnZyOl
1:14pm February 15th 2017 via Hootsuite
Nusbaum: Detection of modified bases - 5-meC, 5-OHmeC, 5-formylcyltosine, 8-oxoguanine. RNA - shows 5 variants at 1% #AGBT17
1:10pm February 15th 2017 via Hootsuite
Nusbaum: 2-3d test by blood culture; need very fast readout, within an hour. #AGBT17
1:09pm February 15th 2017 via Hootsuite
Nusbaum: Speed - often days to get answers, fine. Shows fig from '06 https://t.co/gxQW18lddE surv fx by hour #AGBT17
1:06pm February 15th 2017 via Hootsuite
Nusbaum: Asking for 1 in a million error rate, at a cost of $1,000/sample. Bonus is great reduction in data storage #AGBT17
1:04pm February 15th 2017 via Hootsuite
Nusbaum: Tumor fraction in cfDNA WGS in met cancer: Of 391 pts, 167 from 0.8% to 0.1%; 118 0.1%. Long tail of undetectable #AGBT17
1:03pm February 15th 2017 via Hootsuite
Nusbaum: Why impt? Often you only get one shot - NIPT, cancer ctDNA, monitoring. T21 could be detected in 1955 by cytology. #AGBT17
1:01pm February 15th 2017 via Hootsuite
Nusbaum: Next Accuracy. 'Now this drives me nuts.' 0.5% to 1% since '10. Hasn't progressed at all, oversampling sets a floor on sens #AGBT17
1:00pm February 15th 2017 via Hootsuite
Nusbaum: "I want SVs at the same cost as SNVs, or a small premium." #AGBT17
Nusbaum: Missing data problem - lots of cheap short reads, losing lots of larger variation. Wants structural variation the same way #AGBT17
12:59pm February 15th 2017 via Hootsuite
Nusbaum: Max read length of 150bp, cheap but short. SV's from this '15 TGP paper https://t.co/W3T8jTToyh Need 50kb-500kb cheap #AGBT17
12:58pm February 15th 2017 via Hootsuite
Nusbaum: From $0.10 (to $0.001), for single-cell sequencing. Cheap reads are short - about 150bp. SV graph https://t.co/SaNJBZ3kH1 #AGBT17
12:56pm February 15th 2017 via Hootsuite
Nusbaum: 10K cells? Impractical limits of cost. Asking for a '3-log reduction in the next 10 years'. Sample prep $0.10 for scells #AGBT17
12:54pm February 15th 2017 via Hootsuite
Nusbaum: The cost equation - perhaps hitting a bit of a floor, but it is log-scale. Cost - brain has 10^12 cells; immune is 10^10 #AGBT17
12:53pm February 15th 2017 via Hootsuite
Nusbaum: Not one quest - proposing five quests: cost, length, accuracy, speed, speed, detecting base modifications #AGBT17
12:50pm February 15th 2017 via Hootsuite
Nusbaum: "We are living in the golden age of sequencing". But he is nervous . '03 Collins paper calling for the $1K genome #AGBT17
12:49pm February 15th 2017 via Hootsuite
Chad Nusbaum (Dir Broad Technology MA) The future-history of NGS #AGBT17
12:47pm February 15th 2017 via Hootsuite
Beechem: switch to live feed #AGBT17 and shows sample, and flow cell https://t.co/kRShNoA6fs
12:46pm February 15th 2017 via Hootsuite
Q: Homopolymers? Beechem: For targeted, know what the regions are. Can use blocking oligos to cover these areas. And can stack deck #AGBT17
12:44pm February 15th 2017 via Hootsuite
Beechem: They've been able to do DNA, RNA, short, long-reads, cancer panel, FFPE, no amp, tissue. #AGBT17
12:42pm February 15th 2017 via Hootsuite
Beechem: "I call this fast forwarding" (This is way cool!) Will be talking later with more on the chemistry. #AGBT17
12:40pm February 15th 2017 via Hootsuite
Beechem: Targeted phased long reads: you 'control the cards' - you can look at specific regions that are blocked via hyb to probe #AGBT17
12:39pm February 15th 2017 via Hootsuite
Beechem: Shows ordered reads (simultaneous labels on the same molecule). Able to look at phased results, w/3kb spacing #AGBT17
12:38pm February 15th 2017 via Hootsuite
Beechem: With repeat hexamers, can go to two locations; able to compare DNA and RNA efficiency (very close, RNA is slightly better) #AGBT17
12:36pm February 15th 2017 via Hootsuite
Beechem: Shows data down to 1ng of RNA input. #AGBT17
Beechem: Can sequence RNA and DNA simultaneously (!) Shows DNA and RNA spots. #AGBT17
12:35pm February 15th 2017 via Hootsuite
Beechem: Notes there is no enzymology or polymerase error. The final floor 'may be lower'. Shows RNA direct seq #AGBT17
Beechem: 400 cycles, QV14 at 1-pass, QV22 at 2, QV27 at 3, QV31 at 4-pass, QV 40 at 5-pass. #AGBT17
12:34pm February 15th 2017 via Hootsuite
Beechem: Shows majority vote consensus calling. Shows error statistics - 150K molecules of KRAS, 425 seq probes, 11-plex #AGBT17
12:32pm February 15th 2017 via Hootsuite
RT @ScienceJill: Live-stream of @coregenomics doing full Nanostring seq workflow starting with a single FFPE curl. #agbt17 #thisisreallycool
Beechem: Wrote their own assembler, Pavel Pevzner also wrote one, gets same answers for de-novo 6-mer assembly. #AGBT17
12:31pm February 15th 2017 via Hootsuite
Beechem: #AGBT17 photo of their sequencing Hyb & Seq https://t.co/keAYUFPLuX
12:30pm February 15th 2017 via Hootsuite
Beechem: 512 cards on R1: green-yellow is 'GC'. Next deck is red-red 'TG'. Next is green yellow, 'GC'. 30sec = 'GCTGGC' x 20 million #AGBT17
12:29pm February 15th 2017 via Hootsuite
Beechem: Uniform coverage: want deep, accurate coverage in regions of clinical interest. Thus: can 'stack the deck' to get hi cov #AGBT17
12:28pm February 15th 2017 via Hootsuite
Beechem: "I make the cards, and make up the deck any way I want." Came up with eight decks of cards, 512 reporters at a time #AGBT17
12:27pm February 15th 2017 via Hootsuite
Beechem: Show raw seq images (spots with different colors). With 4096 'cards'; unique hexamers needs to separate out complementaries #AGBT17
12:26pm February 15th 2017 via Hootsuite
Beechem: Hyb of six-base probe, add collection of different probes each with 2-colors, and cycle through the 3 reporters R1-R3 #AGBT17
Beechem: Reading two bases for R1, then two more bases for R2, then two more for R3. Six bases read per cycle. Shows busy slide #AGBT17
12:24pm February 15th 2017 via Hootsuite
Beechem: 4^6 combination of six bases; need 4096 with every base combination. Encode ID of each di-base 'pair'. Two color reporter #AGBT17
12:23pm February 15th 2017 via Hootsuite
Beechem: Reading six bases; last year reported at #AGBT16. Now doing <100bp probes. Detail on bases 1-6, connected to R1-R3 #AGBT17
12:22pm February 15th 2017 via Hootsuite
Beechem: 2000nm large nCounter barcodes; shrunk to 20-50nm in size, cycling reactions 'in seconds' #AGBT17
Beechem: Capture is 5' and 3' bound to flow-cell. Biotinylated on one end, second probe also brought down via Strepavidin/biotin #AGBT17
12:21pm February 15th 2017 via Hootsuite
