Weile: Working with @Invitae looked at their VUS to test, cp their DMS call to indication. 2/2 damaging; 5/5 benign 1 FN, 2 unk #AGBT16
8:19pm February 14th 2017 via Hootsuite
Weile: Describes deep mutational scanning (DMS), mutant ORFs plus barcodes, selection via complementation, then doing the genotyping #AGBT16
8:13pm February 14th 2017 via Hootsuite
Weile: Functional complementation assay - rescuing function by complementation. 'Surprising to see how well it works' #AGBT16
8:10pm February 14th 2017 via Hootsuite
Weile: PolyPhen-2 is popular but to base a clinical decision, precision vs recall curve, only 20% of disease assays would be called #AGBT16
8:09pm February 14th 2017 via Hootsuite
Weile: The problem of VUS: happening fairly often. Everyone has 100-400 rare missense variants never seen before. #AGBT16
8:08pm February 14th 2017 via Hootsuite
Jochen Weile (Uni Toronto CA) Building a functional atlas of all possible mutations in human disease genes. #AGBT16
Adey: sciATAC-Seq, SCI-seq (WGS), sciHiC (C-5), sciRNA-Seq for transcription. (Four papers.) BTW he's looking for a post-doc... #AGBT17
8:05pm February 14th 2017 via Hootsuite
Adey: However even though Li was more complex library, greater bias, impacting copy-number calls. #AGBT17
8:02pm February 14th 2017 via Hootsuite
Adey: Dev model to predict how many unique reads to expect/cell as a function of depth. about 872K reads. Showed Li=more complex #AGBT17
8:01pm February 14th 2017 via Hootsuite
Adey: Two methods - Li salt for disrupt; another is to crosslink and SDS denature. Can get 1200-2200 libraries per PCR plate #AGBT17
7:59pm February 14th 2017 via Hootsuite
Adey: faced challenge of chromatin accessibility. Need to unbind nucleosomes from chromatin w/o disrupting nuclear integrity #AGBT17
7:58pm February 14th 2017 via Hootsuite
Adey: Single-cell barcoding w/o handling or equipment, just a flow cytometer. sciATAC-Seq #AGBT17
Adey: Isolate nuclei, use tranposase-based Index 1 at 96-plex, FACS or dilute to ~20/well. PCR w/index 2. #AGBT17
7:56pm February 14th 2017 via Hootsuite
Adey: Current single cell depends on physical compartments. Their '14 Nature Gen https://t.co/o8VSbsVyZt https://t.co/gJFBgTHJyU #AGBT17
7:55pm February 14th 2017 via Hootsuite
Adey: Application is to look at intra-tumoral heterogeneity. Then tease out drug resistance, metastatic potential etc #AGBT17
7:53pm February 14th 2017 via Hootsuite
Andrew Adey (Oregan Health & Science Univ OR) Sequencing thousands of single-cell genomes with combinatorial indexing #AGBT17
7:51pm February 14th 2017 via Hootsuite
Q: Overcoming RNA degradation from kidney tissue? Bhalla: Most useful to cp groups to groups. 14 types ID'd #AGBT17
7:50pm February 14th 2017 via Hootsuite
Q: 1/10 population, 10K-20K cells/expt? Bhalla: Yes, only did 1,200 cells. #AGBT17
7:48pm February 14th 2017 via Hootsuite
Bhalla: Also could ID a novel cell type (could be transdifferentiation due to Rx, or perhaps new cell types) #AGBT17
7:46pm February 14th 2017 via Hootsuite
Bhalla: Specialized organs with many cell types - can scale to 100's of cells per cell type. Likes the agnostic nature (spp etc) #AGBT17
7:45pm February 14th 2017 via Hootsuite
Bhalla: Fourteen segment kidney, doing PCA, looked at gene expression patterns - very clean separation by PCA #AGBT17
7:42pm February 14th 2017 via Hootsuite
Bhalla: Shows data from '15 paper https://t.co/oozktjQKgN terribly labor-intensive, to look at diuretics, common Rx #AGBT17
7:40pm February 14th 2017 via Hootsuite
Bhalla: Complex and laborious. LCM of 5 nephrons, cut up into sections, and analysis. How it is done today. #AGBT17
7:38pm February 14th 2017 via Hootsuite
Bhalla: So many different cell types along the nephron - 1-2M nephrons within mouse kidney. Tubules, interdigitated cell types #AGBT17
7:37pm February 14th 2017 via Hootsuite
Bhalla: Power of technology - as a nephrologist, to use mouse kidney (not anatomically organized for discovery due to tissue) #AGBT17
7:36pm February 14th 2017 via Hootsuite
Bhalla: RNA in bulk RT'd to cDNA, read barcodes, reconstruct profile of single cells. #AGBT17
7:35pm February 14th 2017 via Hootsuite
Bhalla: Flowcell with 200K wells, with one cell, with one barcoded bead/cell. Lyse, mRNA hyb'd to the bead, then collected #AGBT17
7:33pm February 14th 2017 via Hootsuite
Vivek Bhalla (Stanford Univ CA) Single-Cell RNAseq applications using BD Resolve #AGBT17
7:31pm February 14th 2017 via Hootsuite
Marioni: Acknowledges @coregenomics at the CRUK core facility! #AGBT17
4:58pm February 14th 2017 via Hootsuite
Marioni: Expect work to be published in Science 'in a week's time'. Core activation highly conserved in divergent mice spp #AGBT17
Marioni: Also T-cell activation as a function of aging in mice, comparing gene expression data. #AGBT17
4:57pm February 14th 2017 via Hootsuite
Marioni: Describes work in mouse, CD4+ T-cell activation in vitro, looking at a cre activation pathway between BL6 and CAST lines #AGBT17
4:56pm February 14th 2017 via Hootsuite
Marioni: ...a population but also diff variable genes between populations #AGBT17
4:50pm February 14th 2017 via Hootsuite
Marioni: Shows equations of Poisson dispersion for biological as well a differential comparison. Finding highly-var genes within.. #AGBT17
Marioni: Makes reference to Human Cell Atlas project https://t.co/v650PjQKQz Shows data from https://t.co/4O7uMaldib #AGBT17
4:49pm February 14th 2017 via Hootsuite
John Marioni (CRUK, Univ Cambridge) Aging, evolution and immunity with sc RNA-sequencing #AGBT17
4:46pm February 14th 2017 via Hootsuite
Smet: Patient-derived initiatives like 'Count Me In': Link here: https://t.co/cS21eyVTXp #AGBT17
4:45pm February 14th 2017 via Hootsuite
Smet: 'Walk-up sequencing': pre-made libraries run, whether epigenomics, ChIP-Seq, Hi-C etc. 25% cost reduction for WES #AGBT17
4:44pm February 14th 2017 via Hootsuite
Smet: A 3T flowcell is >250 exomes; higher-plex needs to control for potential contamination #AGBT17
4:43pm February 14th 2017 via Hootsuite
Smet: Cautionary notes: RTA software, integration w/other large-scale projects. Chemistry changes; filling flowcells; batch effects #AGBT17
4:42pm February 14th 2017 via Hootsuite
Smet: Alignment rate high but not as high as the HiSeq 4000; higher # of SNPs and Indels on the NovaSeq 'need further investigation' #AGBT17
Smet: (Only at the Broad, what a place...) :o) #AGBT17
4:41pm February 14th 2017 via Hootsuite
Smet: Showed nice graph of price reduction, including on the X. NovaSeq. "We committed to buying... I can't remember the number" #AGBT17
4:37pm February 14th 2017 via Hootsuite
Smet: (i.e. duplication) - PCR-free started at 15% duplication, down to 3.5%. (What is a 'strip 2'?) #AGBT17
Smet: 38.8K WGS in '16 etc. The HiSeq X 'fully optimized', perfecting PCR-free workflow. Biggest challenge - the 'pad hopping' #AGBT17
4:34pm February 14th 2017 via Hootsuite
Smet: At the Broad: 254K exomes, 67K genomes. 1.6M samples. 26K transcriptomes #AGBT17
4:32pm February 14th 2017 via Hootsuite
Tim De Smet (Broad Inst MA) NovaSeq Broad Institute's comparison and intended uses #AGBT17
4:31pm February 14th 2017 via Hootsuite
Schroth: Cell purity 99.3%; 4922 genes detected; 5.4% ribosomal. Recommend 100K reads/cell. 10B reads = 100K cellsx100K reads #AGBT17
4:28pm February 14th 2017 via Hootsuite
Schroth: Single-cell profiling: solution with Bio-Rad started shipping 'this week'. More detail in the ILMN lounge Wed AM #AGBT17
4:27pm February 14th 2017 via Hootsuite
Schroth: Shows TruSeq DNA methylation 2x75bp PE reads, 6% %Q30 read 1 on NovoSeq. RNASeq data comparable to HiSeq 2500/4000 #AGBT17
4:25pm February 14th 2017 via Hootsuite
