JG:Q:Base modifications, esp w/bacteria? A:Want to look at it (methylation) #AGBT15
3:42pm February 28th 2015 via Hootsuite
JG:Q:Representation bias? A:Homopolymers is an issue, working on self-correction #AGBT15
JG:Q:Did you try Phrap? (Elaine M "You're too young to remember") #AGBT15
3:41pm February 28th 2015 via Hootsuite
RT @mason_lab: #AGBT15 our MinION 2D run from this morning had 13-14% error http://t.co/pj0N8epvP4
3:39pm February 28th 2015 via Hootsuite
JG: 'Future of @nanopore looks pretty bright'. Will be scaling with PromethION. #AGBT15
JG:'I've heard a lot of grumbling about N50 at this meeting'. Graph of completeness of longer elements #AGBT15
3:36pm February 28th 2015 via Hootsuite
JG:Showed bar chart w/ increase in quality. For a N50 of 58kb (short) to 585kb (ONT) #AGBT15
3:35pm February 28th 2015 via Hootsuite
JG: Their GitHub page for NanoCorr: http://t.co/hg82VKi9wU #AGBT15
3:34pm February 28th 2015 via Hootsuite
JG:PacBio error correction algorithms didn't work well for ONT reads; developed their own called NanoCorr #AGBT15
3:33pm February 28th 2015 via Hootsuite
JG:Read once, about 65% identity; so-called 2D base-calling improves to 77% #AGBT15
3:31pm February 28th 2015 via Hootsuite
JG:Yeast W303, best produces 445MB of sequence. 6kb mean readlength. Charted readlength vs % identity #AGBT15
3:30pm February 28th 2015 via Hootsuite
JG:Initial read is k-mers passed through HMM 'can be quite difficult' #AGBT15
3:28pm February 28th 2015 via Hootsuite
James Gurtowski, Cold Spring Harbor Lab. “Error Correction and de novo Assembly of Oxford Nanopore Sequencing” #AGBT15
3:26pm February 28th 2015 via Hootsuite
HJ: Ji Laboratory website at Stanford: http://t.co/KbnxllvTJ7 #AGBT15
3:25pm February 28th 2015 via Hootsuite
HJ:Q:Chromothrypsis? A:May need other types of analysis, suspect clonal in nature #AGBT15
3:24pm February 28th 2015 via Hootsuite
HJ:Q:What orthogonal validation? A:Nuclear trio did @PacBio and single primer extension. FP's were in repetitive regions #AGBT15
HJ: Another example, 17p haplotype and TP53 - one haplotype completely reduced. Yet another: 5q #AGBT15
3:22pm February 28th 2015 via Hootsuite
RT @erlichya: This #AGBT15 meeting is one of the strongest that I can remember in the last few years.
3:18pm February 28th 2015 via Hootsuite
HJ: Showed one example haplotypes being reduced even with 20% normal stroma #AGBT15
HJ: Can use barcode counts as a signal - can use to locate CNV by density along a given chr (i.e. chr8p) #AGBT15
3:17pm February 28th 2015 via Hootsuite
HJ:Phasing metrics: 95%. Copy number plot shows major instability genome-wide. #AGBT15
3:16pm February 28th 2015 via Hootsuite
.@cshperspectives Interesting question, yes term is not ideal. Can't think of better term though at the moment... #AGBT15
3:13pm February 28th 2015 via Hootsuite in reply to
HJ:Used to analyze primary colorectal cancer. Samples are var quality, cellularity, mixture of stroma & normal #AGBT15
3:11pm February 28th 2015 via Hootsuite
HJ: 126K barcodes, 99.0% SNPs phased; longest phased block 11.5MB. Long switch error rate 0.00023 #AGBT15
3:10pm February 28th 2015 via Hootsuite
HJ:Adv is HMW DNA. Var + indexing code. 100K droplet partitions, 100's of thousands of barcodes, 300 genome equivalents #AGBT15
3:09pm February 28th 2015 via Hootsuite
HJ: 'Quantum genetics' - interrogate individual molecules in partitions. Collaboration w/10X Genomics about 1.5y ago #AGBT15
3:08pm February 28th 2015 via Hootsuite
HJ: Implications of phased genomes: cancer rearr, loci w/dense variation (HLA), haplotypes in familial diseases; aneuploidy. #AGBT15
Hanlee Ji (Stanford) "Megabase phasing of genetic aberrations from primary colorectal tumors" #AGBT15
3:07pm February 28th 2015 via Hootsuite
RT @bioinfomagician: Drop-Seq: WANT! #AGBT15 http://t.co/t7der7zMfg
3:05pm February 28th 2015 via Hootsuite
EM:Q:Doublet rates? A:Species mixing expt's very impt. Concentrations adjusted carefully #AGBT15
3:03pm February 28th 2015 via Hootsuite
EM:Q:How are cells handled in device, effect on exp? A:Retina data confirms w/in-situ data #AGBT15
EM:Q:Can I buy one? A:Will provide designs to fabricate devices, company to make beads #AGBT15
3:01pm February 28th 2015 via Hootsuite
EM:Equipment: inverted microscope and 3 syringe pumps. Aspiration: cellular diversity, understand pathological states #AGBT15
3:00pm February 28th 2015 via Hootsuite
EM: ID's known markers of spec amacrine subpopulations. Specs: 10K libraries/day; 12h from cell to sequencer; $0.06/cell. #AGBT15
2:59pm February 28th 2015 via Hootsuite
EM: Separation in 2D is from PCA relative weight. ID 39 retinal types; able to generate dendrogram easily from literature #AGBT15
2:57pm February 28th 2015 via Hootsuite
EM: 570 cells, 10's of thousands of genes detected. Moving to 44.8K retina cells. Found 32 PCA's. Generated movie of a 2D plot #AGBT15
2:56pm February 28th 2015 via Hootsuite
EM: Use a version of SMART (template switch PCR). Used species mixing expts to guarantee each droplet had one cell only in it. #AGBT15
2:53pm February 28th 2015 via Hootsuite
EM:(Reminiscent of RainDance). Movie of droplet generation (nL volumes) Library generation in a single tube of 10K cells #AGBT15
2:52pm February 28th 2015 via Hootsuite
EM: 12 rounds -> 16M barcodes. Several million primers on each bead. Then: generate droplets. Simplest design: oil pinching off #AGBT15
2:51pm February 28th 2015 via Hootsuite
EM: Needed to barcode beads w/PCR barcode, cell barcode, UMI. Did split-and-pool oligo synthesis #AGBT15
2:50pm February 28th 2015 via Hootsuite
EM:Describes emulsions - shows monodisperse droplets. Concept: lyse in emulsion, beads bind to RNA in bubble #AGBT15
2:49pm February 28th 2015 via Hootsuite
EM: Wish list for single-cell: scale (10^4-10^5 cells); flexible #'s (10 or 10K); sizes (2um - 50um); easy to use #AGBT15
2:48pm February 28th 2015 via Hootsuite
EM: Summary chart of # cells vs # of genes. CYTOF on far right, ISH toward 1/1, C1 / plate scRNA-Seq. Chart of cost for 500 cells #AGBT15
2:47pm February 28th 2015 via Hootsuite
EM: In the brain an est >1K cell subtypes. Study cell types & states in complex tissues? Illus bulk with smoothie / fruit salad #AGBT
2:46pm February 28th 2015 via Hootsuite
Evan Macosko, Harvard Med. DropSeq: A Droplet-Based Technology for Single-Cell mRNA-Seq Analysis on a Massive Scale #AGBT15
2:45pm February 28th 2015 via Hootsuite
.@CLCbio Many thanks! I did stop by last night! :) #AGBT15
2:42pm February 28th 2015 via Hootsuite in reply to
MT: CRISPR/Cas9 in human cells 'highly spec'; vary by gRNA design; full spectrum of var is inducible; NAHR-med disorders modeled #AGBT15
2:41pm February 28th 2015 via Hootsuite
MT: Designed an RNA-seq experiment to validate; data shows dups increased; deletion decreased. #AGBT15
2:40pm February 28th 2015 via Hootsuite
MT: Separated single iPSC via isolation by FACS; iPS cells. Method made both microdels and seg duplications. #AGBT15
2:38pm February 28th 2015 via Hootsuite
MT: Designed unique single gRNAs that can target each 16p segmented duplication. Ended up w/one copy of gene #AGBT15
2:37pm February 28th 2015 via Hootsuite
