Q: ASPIRE starting material? Time? Patel: Purified DNA. Time hasn't been optimized. #Tricon
5:45pm February 22nd 2017 via Hootsuite
Patel: Working on more complex samples; can sequence DNA and RNA directly. Methods dev to deconvolute messy data #Tricon
5:44pm February 22nd 2017 via Hootsuite
Patel: ONT is constantly improving; 2nd strand can be lower quality; 1D squared may improve. Selective seq: Cas9-based selection #Tricon
5:43pm February 22nd 2017 via Hootsuite
Patel: Higher stringency, fewer reads. Can imagine bacterial DNA and host DNA where the host DNA is rejected #Tricon
5:40pm February 22nd 2017 via Hootsuite
Patel: Since read-until is in real-time, able to get higher rDNA mappability depending on level of stringency #Tricon
5:39pm February 22nd 2017 via Hootsuite
Patel: Modified version - a local base-calling to match against reference. Did Cas9-based selection of rDNA, enrich for spec size #Tricon
Kirshnakumar: Read-until allows for pruning of data in real-time. '16 Nature Methods https://t.co/nnG8lmStyC Can reject reads #Tricon
5:36pm February 22nd 2017 via Hootsuite
Kirshnakumar: Then looked at Kmer calls - Shows a plot of all 1024 5-mer possible calls, and expected vs observed ratios #Tricon
5:29pm February 22nd 2017 via Hootsuite
Kirshnakumar: The % that the next base will be similar to the score of that base; but remember, these are 5-mer calls. #Tricon
5:28pm February 22nd 2017 via Hootsuite
Kirshnakumar: Looked at contiguous base calls and Qscores; binned into deciles, shows base calling mistakes are non-deterministic #Tricon
5:27pm February 22nd 2017 via Hootsuite
Kirshnakumar: QC span, shows violin plots of GC content (B.thail 68%, C.diff 28%, E.coli 51%). #Tricon
Kirshnakumar: Population of reads <1kb w/low Qscores; above 1kb a cloud, scores range from 7-15. #Tricon
5:26pm February 22nd 2017 via Hootsuite
Kirshnakumar: Get 80-90% identity to the genomes in terms of alignment. 2D histograms show Qscore not related to read length #Tricon
5:25pm February 22nd 2017 via Hootsuite
Kirshnakumar: For Dx, Poreminion, Poretools, in-house scrips. 3 orgnsms, E.coli, C.diff., B.thail. Pie charts of 1D Q>6, 2D Q>9 #Trico
5:24pm February 22nd 2017 via Hootsuite
Kirshnakumar: Raw squiggle data (Metrichor, Nanonet is @oxfordnanopore callers. Mapping via LAST, BWA-MEM, GraphMap #Tricon
5:23pm February 22nd 2017 via Hootsuite
Kirshnakumar: Ideal for rapid bacterial ID in the field, wanted to develop an end-to-end pipeline. #Tricon
5:22pm February 22nd 2017 via Hootsuite
Kirshnakumar: Windows are called as 5-mers; oppy to use on both template and complement ("2D") #Tricon
5:21pm February 22nd 2017 via Hootsuite
Kirshnakumar: Data is different; shows squiggle plot, and neural network basecallers looking at 5-mers. #Tricon
Kirshnakumar: Field sequencing already - ISS, Ebola, ZiBRA (Zika virus in razil) #Tricon
5:20pm February 22nd 2017 via Hootsuite
Kirshnakumar: Very interested in its long reads, real-time data, interactive 'read-until', field-read and small footprint #Tricon
5:19pm February 22nd 2017 via Hootsuite
Raga Kirshnakumar (Sandia NL): Real-time diagnostics using nanopore sequencing #Tricon
Patel: Miniaturization of rotary pump, actuator; auto placement of magnet; perhaps lyophilization. Wants to include DNA extraction #Tricon
5:18pm February 22nd 2017 via Hootsuite
Patel: Volume of reagents are the same as in bench protocol. Can be hands-off; load from capillary to load. #Tricon
5:17pm February 22nd 2017 via Hootsuite
Patel: Can do automated 2D prep 'for field-portable sequencers'. Simple to make, disposable components, semi-automated #Tricon
5:16pm February 22nd 2017 via Hootsuite
Patel: Showing E coli data, size distribution of reads looks comparable; mean RL 4.4kb vs 4.5kb manual. Alignment 85% vs 88% #Tricon
Patel: Has an automated library prep platform, showed data with 48kb library prep vs manual; DNA yield comparable #Tricon
5:14pm February 22nd 2017 via Hootsuite
Kamlesh Patel (Sandia NL): Portable sequencing: fundamentals in seq technology with nanopores, sample prep and data analysis #Tricon
5:13pm February 22nd 2017 via Hootsuite
Q: source of noise in ddPCR? Singh: some get mutant and WT both templates are there #Tricon
3:08pm February 22nd 2017 via Hootsuite
Singh: Points out the importance of using controls - (SeraCare has them.) https://t.co/jeMAm7uG3w #Tricon
3:07pm February 22nd 2017 via Hootsuite
Singh: They can use cell-culture DNA as a surrogate. #Tricon
3:05pm February 22nd 2017 via Hootsuite
Singh: Showed LoD studies, down to 0.3%; blow that was not consistent. 3 droplets was their cutoff #Tricon
Singh: RET M918T test is via ddPCR; showed validation cohort n=63. They use 10ng input; cp ccfDNA and FFPE. R^2=9.94 Pyro vs dPCR #Tricon
3:04pm February 22nd 2017 via Hootsuite
Singh: Test for melanoma, lung ca, medullary thryroid for ctDNA #Tricon
3:01pm February 22nd 2017 via Hootsuite
Singh: Shows scan of plasma ccfDNA and FFPE, with 140bp peak. Are not enthusiastic about synthetics. #Tricon
3:00pm February 22nd 2017 via Hootsuite
Singh: ccfDNA technology choices paper '15 PLoS https://t.co/HPAVJTlRwA #Tricon
2:58pm February 22nd 2017 via Hootsuite
Singh: Did cp of QIASymphony vs manual of 3mL plasma; showed comparable yield and Alu115/Alu247 also very similar #Tricon
2:57pm February 22nd 2017 via Hootsuite
Singh: Purple tops are oka, but only within 16h; shows Mehrotra JMD (under review) data #Tricon
2:56pm February 22nd 2017 via Hootsuite
Singh: For testing platforms - Allele-spec qPCR/ddPCR/NGS/BEAMing? DNA input? workflow, sensitivity? #Tricon
2:54pm February 22nd 2017 via Hootsuite
Singh: We are taking NGS tumor assays that were 5% or 2% and pushing the limit. Logistics - what to isolate? How? #Tricon
2:53pm February 22nd 2017 via Hootsuite
Rajesh Singh (MD Anderson TX) Mutation screening of liquid biopsies: technical challenges. #Tricon
2:51pm February 22nd 2017 via Hootsuite
Luthra: Challenges - lack of harmonization for pre-analytic steps such as peri blood collection, storage, ctDNA isolation #Tricon
2:50pm February 22nd 2017 via Hootsuite
Luthra: Feb 17 single-gene test by ddPCR, developed in-house. Has collaboration with Guardant Health, co-develop 73 cancer genes #Tricon
2:49pm February 22nd 2017 via Hootsuite
Luthra: Pt w/colon ca, adjuvant Rx, relapse 1y, plasma at 14mo 39% in plasma but 0% PIK3CA tissue, clonal divergence? #Tricon
2:47pm February 22nd 2017 via Hootsuite
Luthra:.MDACC case studies; Stg IV Colon ca; deep seq w/ Ampliseq; chemo. Found 1% plasma, not in tissue, later found 0.2% in tissue #Tricon
2:46pm February 22nd 2017 via Hootsuite
Luthra: Looking at RET M918T in medullary thyroid carcinoma: If >1%, poor survival. (Master's degree student project) #Tricon
2:44pm February 22nd 2017 via Hootsuite
Luthra:Correlation of BRAF with clinical outcome: BRAFi and MEKi '16 ref https://t.co/TlubGIRTth Looked at BRAF V600E #Tricon
2:43pm February 22nd 2017 via Hootsuite
Luthra: And concordance bet tissue and ctDNA is 32% (panc) to 97% (colorectal). #Tricon
2:39pm February 22nd 2017 via Hootsuite
Luthra: ccfDNA is from both normal and tumor cells; shows long list of 9 references with concordance across cancer types, genes, ct #Tricon
2:38pm February 22nd 2017 via Hootsuite
Luthra: ctDNA discovered in 1947; but decades to develop. Applications for monitoring, MRD, resistance det, tumor burden #Tricon
2:36pm February 22nd 2017 via Hootsuite
Raja Luthra (MD Anderson TX) Mutation screening of liquid biopsies: promise and clinical utility #Tricon
2:34pm February 22nd 2017 via Hootsuite
