Jaffe: 84% of 5+kb reads have at least 1 perf. 50-mer. 100% have at least 1 perf. 25-mer. "Perfect windows punctuated by artifacts" #AGBT14
11:21am February 14th 2014 via Hootsuite
Jaffe: Long tail - histogram went to 12kb, still reads out further. #AGBT14
11:20am February 14th 2014 via Hootsuite
Jaffe: E. coli (6x), Scardovia(13x), 5kb read-length. Mode - about 4.5kb. Measured bioanalyzer against readlengths and it matches #AGBT14
Jaffe: 1 cm2 rect array, 512 pores, speed thru pore configurable. 1 to 100 bases / sec. Data today is 25 bases / sec #AGBT14
11:19am February 14th 2014 via Hootsuite
Jaffe: A probabilistic assignment of the signal. Possible to get the wrong k-mer. Illustrated MinION device, Oreo (TM) shown for ref #AGBT14
11:18am February 14th 2014 via Hootsuite
Jaffe: Each horizontal event turned into a k-mer. Computational model is not trivial - a lookup of an approx table, not discrete. #AGBT14
11:17am February 14th 2014 via Hootsuite
Jaffe: has pA on Y, time in s on X; six-mers along x-axis shown ("CTGCTT" etc.) #AGBT14
11:16am February 14th 2014 via Hootsuite
Jaffe: Current across membrane, >1 base at a time; current a fn of several k-mer bases. Pore -> signal -> kmers -> bases. Raw da
Jaffe: Start with HMW DNA; 'tether' an adapter; illustrating ONT nanopore illustration. Ratchets bases on top of the nanopore #AGBT14
11:15am February 14th 2014 via Hootsuite
Jaffe: Deamer nanopore reference from 1999: http://t.co/CrSPIH090n #AGBT14
11:14am February 14th 2014 via Hootsuite
Jaffe: From 1996 reference: Nanopore "In principle". Will talk about ONT (Oxford Nanopore) #AGBT14
11:13am February 14th 2014 via Hootsuite
Up next: David Jaffe, Broad Institute of MIT and Harvard: Assembly of Bacterial Genomes Using Long Nanopore Reads #AGBT14
11:12am February 14th 2014 via Hootsuite
Schuster: Q: Cross-breed with Andean condor? A: No, it would create a 'genetic freak' #AGBT14
11:11am February 14th 2014 via Hootsuite
Schuster: Informed breeding reference Pac Symp Biocomput 2010 http://t.co/9dFJivqO29 #AGBT14
11:10am February 14th 2014 via Hootsuite
Schuster: Tasmanian devil effort ("informed breeding") now in-place at the SD Wild Animal Park condor breeding facility #AGBT14
11:06am February 14th 2014 via Hootsuite
Schuster: Now searching for the 'super condor' in museum specimens. Then re-insert the alleles to strengthen existing species #AGBT14
11:05am February 14th 2014 via Hootsuite
Schuster: First time parthenogenesis shown in birds. 1345 SNPs to look at maternal homozygous calls. 1.4% errors or misaligns #AGBT14
11:04am February 14th 2014 via Hootsuite
Schuster: Parthenogenesis - of 625 birds, 2 parthenotes were discovered ("no contribution of the father") 41 SNPs in CFTR #AGBT14
11:03am February 14th 2014 via Hootsuite
Schuster: The big question: what kind of diversity? mtDNA 'looks grim': only 4 haplogroups. #AGBT14
11:00am February 14th 2014 via Hootsuite
Schuster: Assy 2 to be released later this year, N50 68kb scaffolds. Illustrated Moleculo data length distribution (longest 27kb) #AGBT14
10:56am February 14th 2014 via Hootsuite
Schuster: Uses a combination for genome assy: MiSeq, stitched to make longer 454-style, HiSeq, Moleculo & 454 data #AGBT14
10:55am February 14th 2014 via Hootsuite
Up next: Stephan Schuster (Penn State): California Condor: Population Genomics and Informed Breeding Aid Species Conservation #AGBT14
10:54am February 14th 2014 via Hootsuite
Suzuki: Q:(Mardis) What's the time-factors involved? A: 1w to 2w is reasonable timeframe. #AGBT14
10:18am February 14th 2014 via Hootsuite
Re: DeRisi talk - no focus on speeding up library prep or seq, only 90m of bifx. IMHO 2 PGMs, 318 chip = 8M reads in ~7h, not o/n #AGBT14
10:16am February 14th 2014 via Hootsuite
Suzuki: From 116 samples, looking at genotype in both host / parasite and gene exp. in host / parasite #AGBT14
10:10am February 14th 2014 via Hootsuite
Suzuki: Plasmodium tags vs. human - both in vitro and experimental correlated well. 10% of reads were Pf (Plasmodium f.) #AGBT14
10:07am February 14th 2014 via Hootsuite
Suzuki: 'Interactive' Transcriptome analysis - parasites mixed with host human cells, able to monitor both human and plasmodium exp. #AGBT14
10:06am February 14th 2014 via Hootsuite
Suzuki: Unculturable, focused in on collab. with Manado Indonesia (Coelacanth captured there) #AGBT14
10:05am February 14th 2014 via Hootsuite
Up next: Yutaka Suzuki, University of Tokyo. Interactive Transcriptome Analysis of Malaria Patients #AGBT14
10:04am February 14th 2014 via Hootsuite
DeRisi: Q: Just bulk library prep and seq? A: If you know what's wanted, but unknown if virus, bact. Just total nucl. acid #AGBT14
10:03am February 14th 2014 via Hootsuite
DeRisi:Q:(Mardis): Any way to systematize this? A: Autoimmune test on rat brain slices, proteomics etc. Also seq 'everything - done' #AGBT14
10:01am February 14th 2014 via Hootsuite
DeRisi: Q:(Mardis) Any other poss. other than brain biopsy? A: Seq serum as well, no reads there. #AGBT14
10:00am February 14th 2014 via Hootsuite
DeRisi: So the regular Leptosp. test came back negative, and would have come back in 2w - hard to culture. #AGBT14
9:59am February 14th 2014 via Hootsuite
DeRisi: Further work - confirmed strain to Puerto Rico. But published CDC test for validated test is only 60% sensitive. #AGBT14
9:58am February 14th 2014 via Hootsuite
DeRisi: IRB did not allow treatment decision. Bioethicist, trtmnt. was penicillin in this case. High dose IV, 14d later discharged! #AGBT14
9:57am February 14th 2014 via Hootsuite
DeRisi: Likely got it in PR (Puerto Rico). Only 100-200/y in US, meningitis in ~15% of cases. #AGBT14
9:56am February 14th 2014 via Hootsuite
DeRisi: Find Leptospiria. Breakdown of viral and bacterial reads. 475 reads, distributed all over Leptospiria. Widespread zoonosis #AGBT14
9:55am February 14th 2014 via Hootsuite
DeRisi: Now goes to uBiome browser (UCSF Kickstarter project) produced a taxonomy browser UCSF press rel: http://t.co/aD9ULWCoSb #AGBT14
9:54am February 14th 2014 via Hootsuite
DeRisi: Breakdown of 90m of analysis, this is AGBT of course. 8.1M reads, used SNAP, 65m took non-human reads #AGBT14
9:52am February 14th 2014 via Hootsuite
DeRisi: Rapid IRB at day 42: "How fast can you do it?" Answer: 2 days on MiSeq. Metagenomic prep, 1d library, o/n run, 90m analysis #AGBT14
9:51am February 14th 2014 via Hootsuite
DeRisi: YouTube video screenshot from mother (not shown), 22 pumps going into the boy. Lab reached out to then. #AGBT14
9:50am February 14th 2014 via Hootsuite
DeRisi: Now at a critical point, put on steriods, continued decline, put into an induced coma due to seizures. #AGBT14
9:49am February 14th 2014 via Hootsuite
DeRisi:Encephalitis - could be infection or autoimmunity, with drastic treatment differences in this case. 1cm2 frontal lobe biopsy #AGBT14
9:47am February 14th 2014 via Hootsuite
DeRisi: Infectious disease workup - almost 30 of them - all negative. #AGBT14
9:46am February 14th 2014 via Hootsuite
DeRisi: Jul '13 boy had abnormal MRI - 'hyperintensity' images doesn't indicate cause. #AGBT14
9:45am February 14th 2014 via Hootsuite
DeRisi: Sharing story of 14y.o. WI boy, lot of symptoms from Sep '12 through Jul '13; exposure Aug '12 'at a FL resort' or PR visit #AGBT14
.@JChrisPires Thanks for the shoutout!
9:42am February 14th 2014 via Hootsuite in reply to
DeRisi: NGS for ICU / critical care cases? Cons are speed (incl. analysis) and cost #AGBT14
9:40am February 14th 2014 via Hootsuite
DeRisi: In CA: 7y study, 1570 cases; 63% undiag., as it can be caused by >100 pathogens (could be autoimmune also). #AGBT14
9:39am February 14th 2014 via Hootsuite
DeRisi: 20K cases/y in US, 5.8% fatal, 50% undiag. (!) Cost in '10 ($2B in US); long hospital stays. #AGBT14
9:38am February 14th 2014 via Hootsuite
