Bustamante: 43k pre-capture ILMN lib. SNPs, 400K variable sites on Proton, the PCA looks much better. Several 2012 papers #AGBT14
11:11am February 15th 2014 via Hootsuite
Bustamante: Otzi 'iceman' story: the PCA with Sardinians - 5.3k y.o. border of Italy, was he on vacation in the Alps when he died? #AGBT14
11:10am February 15th 2014 via Hootsuite
Bustamante: Proton data allows them to find ground-truth data for ancient samples. DNA damage evidence toward beg. and end #AGBT14
11:09am February 15th 2014 via Hootsuite
Bustamante: Showed some @IonTorrent Proton data, in collaboration with T Harkins. Pointed out the help of orthogonal technology #AGBT14
11:08am February 15th 2014 via Hootsuite
Bustamante: Get 15-60% of reads to map to the human genome; AJHG 2013 paper #AGBT14
Bustamante: Use biotinylated RNA baits to pulldown whole-genome human 1% from an environmental paleo sample #AGBT14
11:02am February 15th 2014 via Hootsuite
Ehrich: Q: Is there latitude for consent? A: Existing regulations - OK for QC, also specialized consent post-test #AGBT14
10:30am February 15th 2014 via Hootsuite
Ehrich: Q: Perturbations - systemic change to reagents? A: Chem versions is a big deal, several $MM. All the revalidation... #AGBT14
10:29am February 15th 2014 via Hootsuite
Ehrich: Q: Neutrophil extracellular traps of cell-free DNA? A: No, b/c of QC in-place. No a priori exclusion criteria. >4% fetal: OK #AGB
10:28am February 15th 2014 via Hootsuite
Ehrich: Q: How to detect bal tranloc? A: Map into the read. Unperfect matches can have a linear model applied #AGBT14
10:27am February 15th 2014 via Hootsuite
Ehrich: Conclude: How do we relay info back? 4M preg's/y (US); 3-5% of these have a genetic condition. More genetic counseling too #AGBT14
10:26am February 15th 2014 via Hootsuite
Ehrich: Have identified balanced translocations - 'but no one other than us seem to care'. Their algorithm is single-base resolution #AGBT14
10:22am February 15th 2014 via Hootsuite
Ehrich: Can find 'outside prenatal' - ID 'uterine ploemorphic leiomyosarcoma'; some will have disease DNA in the sample #AGBT14
10:21am February 15th 2014 via Hootsuite
Ehrich: Can find down to 3-6MB, but only 60-85% sens, still >99% spec. Better at >7MB #AGBT14
10:19am February 15th 2014 via Hootsuite
Ehrich: Ability to detect depends on: size of indel; counts; fetal fx.; sequence noise. Plan to work with existing reads / process #AGBT14
10:18am February 15th 2014 via Hootsuite
Ehrich: Now looking at sub-chromosomal rearrangements. Now a signal-to-noise deviation problem, with subsampled reads #AGBT14
10:17am February 15th 2014 via Hootsuite
Ehrich: Have found instances of confined placental mosaicism; co-twin demise; fetal mosaicism; maternal mosaicism or malignancy #AGBT14
10:15am February 15th 2014 via Hootsuite
Ehrich: Only offered to high-risk pts.; 2.14% positive rate identical to expected incidence rate #AGBT14
10:14am February 15th 2014 via Hootsuite
Ehrich: 'no result' in 0.8% (insuff. DNA); 'no result' 0.3% (technical) Ave gestational age 14w 6d TA 4d 3.7 multi-fetal rate #AGBT14
10:13am February 15th 2014 via Hootsuite
Ehrich: Trisomy 21 (T21) test lists several 2011 - 2013 clinical validation data publications. >200K samples tested; 150k in '13 #AGBT14
10:12am February 15th 2014 via Hootsuite
Ehrich: NGS use is only chromosomal copy number of fetal origin, 36-bp. #AGBT14
10:05am February 15th 2014 via Hootsuite
Ehrich: 90% of free DNA is from mother. Amt of Chr21 is only 1.35% of the fetal genome. #AGBT14
10:04am February 15th 2014 via Hootsuite
Ehrich: 97.7% of amniocentesis cases are normal. Lo discovery in 1997 (Lancet) of cell-free fetal DNA http://t.co/1QKqUhiC9T #AGBT14
10:03am February 15th 2014 via Hootsuite
Up next: Mathias Ehrich, Sequenom. “Non Invasive Prenatal Testing: A Showcase of High Volume Next Generation” #AGBT14
9:59am February 15th 2014 via Hootsuite
Blainey: Q: When for sale? A: We have no interest in commercializing. #AGBT14
Blainey: Q: For metagenomics, lysis can be a challenge to get 100% DNA downstream A: Acknowl. the problem, not easy to solve #AGBT14
9:58am February 15th 2014 via Hootsuite
Blainey: A: (con't) They are doing short-read libraries, using Tn5-based fragmentation, all integrated but hasn't been done yet #AGBT14
9:57am February 15th 2014 via Hootsuite
Blainey: Q: How will it handle 200kb DNA? A: Believes it will work well. Low Reynolds number. #AGBT14
9:56am February 15th 2014 via Hootsuite
Blainey: Cost calcs: 128/run, $5K controller, reagents list is $0.94. FWIW $0.01/sample for the amortized cost of a pipettor. Hmm. #AGBT14
9:54am February 15th 2014 via Hootsuite
Blainey: Showing quality and contam. levels, a continuous flow device, 'libraries/hour'. 1 chip = 1k libraries/day (theoretical) #AGBT14
9:53am February 15th 2014 via Hootsuite
Blainey: Not sure if this is the Quake ref from 2005 for library prep purpose, but here you go. http://t.co/lYmO62BWTt #AGBT14
9:52am February 15th 2014 via Hootsuite
Blainey: 'Chips have a productive lifetetime of months' #AGBT14
9:49am February 15th 2014 via Hootsuite
Blainey: Tech called 'sieve valve' from Stephen Quake 2005, Tan et al PLoS Ref http://t.co/xWmcvhI8yV #AGBT14
9:48am February 15th 2014 via Hootsuite
Blainey: Micron-scale particles for capture, solid-phase reversible immob. tech (ChargeSwitch anyone?) elute, flush, mix. #AGBT14
9:46am February 15th 2014 via Hootsuite
Blainey: $1 = 10 pipet tips. So they have to think re-use, fully integrated, low capital, HTP, output already pooled #AGBT14
Blainey: Also HTP and the macro to miniaturized world interface. Set goal: library prep for $1 (typically about $50 to $120 now) #AGBT14
9:44am February 15th 2014 via Hootsuite
Blainey: Illustrates the miniaturization in fluidics and how it can help lower costs. Fast TAT, low-input; sample tracking #AGBT14
9:43am February 15th 2014 via Hootsuite
Blainey: Illustrates the inherent disconnect between per-sample library cost (esp for indiv. bacterial cells, also exomes) cp to NGS #AGBT14
9:41am February 15th 2014 via Hootsuite
Blainey: Feels that sample prep library construction technology is becoming an important area for development #AGBT14
9:36am February 15th 2014 via Hootsuite
Up next: Paul Blainey, Broad Institute of MIT and Harvard: Highly Integrated Microfluidic Shotgun Library Construction #AGBT14
9:35am February 15th 2014 via Hootsuite
Schloss: Showed slide from 2011 NGHRI 10y strategic plan, 'Base Pairs to Bedside' (PDF) http://t.co/jl0P5SqvPJ #AGBT14
9:34am February 15th 2014 via Hootsuite
Schloss: "Quantum tunneling" is something to keep an eye on, IMHO. #AGBT14
9:32am February 15th 2014 via Hootsuite
Schloss: NHGRI commitment; Inst. peer-review; advisors supportive; discovery/dev./investment chain. Diversity & competition #AGBT14
9:31am February 15th 2014 via Hootsuite
Schloss: Why does this work? Grantees share their work; motivated by meetings like this one; compete & collab.; students come #AGBT14
9:30am February 15th 2014 via Hootsuite
Schloss: (On that - take a look at this site - http://t.co/aJNWIygjlW Spoke to CEO this AM, they announced 22-mer seq recently.) #AGBT14
9:29am February 15th 2014 via Hootsuite
Schloss: Alpha hemolysin pore, (ONT), tour of singe graphene nanopore and next quantum tunnerling #AGBT14
9:28am February 15th 2014 via Hootsuite
Schloss: Motivations for single molecule: Direct read; long read; non-destructive, modified bases, RNA? digital readout, easy prep #AGBT14
9:26am February 15th 2014 via Hootsuite
Schloss: PacBio tech. illustrated (Korlach & Turner);'96 PNAS Deamer ref. - concept of nanopores. #AGBT14
9:25am February 15th 2014 via Hootsuite
Schloss: Compares 3 mo, 100 CE machines in '03, now 1 day 1 machine '14. Shows 'classic' NGHRI slide here: http://t.co/AsWXuJmz5F #AGBT14
9:23am February 15th 2014 via Hootsuite
Schloss: (Ah the memories...) Other systems - microfluidic sample prep for DNA sequencing; ALL at Duke (acq. by ILMN) #AGBT14
9:21am February 15th 2014 via Hootsuite
