FK:Protein was Retroviral protease in monomeric form. (Mason-Pfizer Monkey Virus). 3 players: credited their teams in publication! #ASHG14
1:48pm October 19th 2014 via Hootsuite
FK: In 5d, 2,771 model is close, shared to others, second gamer improved, 3rd player then solved it. Drastic change over starting #ASHG14
1:47pm October 19th 2014 via Hootsuite
FK: Cp of performance presented, best pub > gamer > unpubl method. Next: unsolved structure to gamers #ASHG14
1:46pm October 19th 2014 via Hootsuite
FK: Example of Blue Fuse (top player algorithm) wrote a script, very close to the lab postdoc's method. #ASHG14
1:45pm October 19th 2014 via Hootsuite
FK: Remarkable that players share with their competition. A cookbook - and stats on player behavior during CASP9. #ASHG14
1:43pm October 19th 2014 via Hootsuite
FK: Gamers outperforming in 2 cases the state-of-the-art computational models. Player strategy wiki - sharing best recipes #ASHG14
1:42pm October 19th 2014 via Hootsuite
FK: CASP experiment in the real-world (computational biol effort). 'Foldit players (since '10) had most success in Refinement' #ASHG14
1:41pm October 19th 2014 via Hootsuite
FK: Human players go in the wrong direction (energy-wise for protein folding) but end up getting much better to optimum. "Foresight" #ASHG14
1:40pm October 19th 2014 via Hootsuite
FK: Large % of Foldit players have no protein or other science experience. Looking at player solutions cp to computers #ASHG14
1:39pm October 19th 2014 via Hootsuite
Firas Khatib (UMass Dartmouth) Crowd computing: Scientific discoveries by protein folding game #ASHG14
1:37pm October 19th 2014 via Hootsuite
YE: 'Just the first step' More analysis of 23andMe data http://t.co/a1gjJqIqCt User:demo pass:Demo 'We'd love to hear your feedback' #ASHG14
1:25pm October 19th 2014 via Twitter Web Client
YE: Data release: de-ID data at FamiliLinx http://t.co/RCkD1i6pHm now available for use #ASHG14
1:21pm October 19th 2014 via Hootsuite
YE: Their data hints that longetivity is additive. Looking further: monozygotic twins, 934 MZ from Danish Twin Registry. Confirmed #ASHG14
1:20pm October 19th 2014 via Hootsuite
YE: Families are complicated by consanguinity. Looking at IBD, fitting 2.1M datapoints, IBD x Correlation, comparing additive vs epi #ASHG14
1:19pm October 19th 2014 via Hootsuite
YE: Adjusted dataset for 3 major wars and sex differences; looking at heritability for parent/child lifespan. (h^2 = 17%) #ASHG14
1:17pm October 19th 2014 via Hootsuite
YE: Impact of WWI/II, and impact of improved public health and medical care visuallized #ASHG14
1:15pm October 19th 2014 via Hootsuite
YE: Validation data of age of death distribution shown, death in infancy has high errors (expected) Year by age of death shown too #ASHG14
YE: Phenotype of longetivity: easy to measure, important trait. GWAS not successful Sebastiani 2013 http://t.co/JPQ9aOASS0 #ASHG14
1:14pm October 19th 2014 via Hootsuite
YE: Movie of settlement shown. http://t.co/ZvwofvnoEH website: http://t.co/HmV8MJcNsf Plot of migration, distance between relatives #ASHG14
1:12pm October 19th 2014 via Hootsuite
YE: Ex Boston in 1650, lines are very close. Sydney at 1800 etc. Some shift but still very close #ASHG14
1:11pm October 19th 2014 via Hootsuite
YE: Result: World map of density. But is it correct? Looked at 10 cities, over starting dates, cross-matched with 'first settlement' #ASHG14
1:10pm October 19th 2014 via Hootsuite
YE: 24 steps from Wright... Events of birth, residences, death, burial. How to convert free text to GPS? Yahoo Geoparser #ASHG14
1:09pm October 19th 2014 via Hootsuite
YE: Largest family tree has 13M individuals. Plot of 70K (0.5% of data) - his computer doesn't have enough pixels to visualize (!) #ASHG14
1:08pm October 19th 2014 via Hootsuite
YE: 43.7M public profiles using 'Geni' API, >1M geneologists, some contributed 100K profiles. 50% data from a few thousand #ASHG14
1:07pm October 19th 2014 via Hootsuite
YE: Traditionally - took 5y to get 200 second cousins. So: used social media instead. Filled out tree from a 3rd cousin online #ASHG14
1:06pm October 19th 2014 via Hootsuite
YE: Debate betw. Fischer (Falconer 1988) vs Wright, a long debate. Missing heritability - need for large kinships to separate curves #ASHG14
1:05pm October 19th 2014 via Hootsuite
YE: Complexity of genetics - additive vs. epistasis models is compared. Emph. polynomial function, a complex, exponential curve #ASHG14
1:04pm October 19th 2014 via Hootsuite
YE: iPipet published in Nature Methods here http://t.co/lJikrEPOhN #ASHG14
1:03pm October 19th 2014 via Hootsuite
YE: A few years ago published this in Science http://t.co/tltzY8lJNb now http://t.co/EFAPR9zm8a for $300, can be shared online #ASHG14
1:02pm October 19th 2014 via Hootsuite
Session: Crowdsourced Genetics; Y Erlich @erlichya Dissecting the genetic architecture of complex traits with millions of people #ASHG14
1:00pm October 19th 2014 via Hootsuite
.@omespeak Atul is @atulbutte , not sure about the others. #ASHG14
1:00pm October 19th 2014 via Hootsuite in reply to
MK:Q:How to cleanup public DB's? (Q from Twitter) A:NHANES is very clean data; others not so. Sharing movement will help. #ASHG14
12:35pm October 19th 2014 via Hootsuite
MK:Q:Payment issues? A:Ins. Co's don't 'believe clin. utility has been established'. BRCA testing took 11y before it was a guideline ASHG14
12:33pm October 19th 2014 via Hootsuite
MK: Need strong epidem. foundation, knowledge integration; genomics data promising, but cautions against overpromising #ASHG14
12:31pm October 19th 2014 via Hootsuite
MK: Covering EGAPP Working Group. The evidence gaps exists, 10 rec's to-date, 3 tiers website: http://t.co/OSv1VybIqZ #ASHG14
12:28pm October 19th 2014 via Hootsuite
MK: Points to Schully et al 2014 http://t.co/497ViwQSiS on evidence synthesis guiding genomic medicine #ASHG14
12:26pm October 19th 2014 via Hootsuite
MK: 'Watson's going to need humans to input the data' #ASHG14
12:24pm October 19th 2014 via Hootsuite
MK: Brought up figure from this recent publication, the 'interpretation bottleneck for pers. med.' http://t.co/MV5M0QP1BJ #ASHG14
MK:Working with the NCI Cohort Consortium, going beyond gene discovery. Data harmonization and data sharing #ASHG14
12:21pm October 19th 2014 via Hootsuite
MK: 4 things they are doing downstream of discovery: evaluation, implementation, effectiveness, feedback back to implementation #ASHG14
12:20pm October 19th 2014 via Hootsuite
MK: (Was able to locate the cartoon here: http://t.co/iTQsAqGUKg ) #ASHG14
12:17pm October 19th 2014 via Hootsuite
MK: Cartoon: "Today's random medical news" with 'spin the wheel, from the Cinncinnati Enquirer #ASHG14
12:15pm October 19th 2014 via Hootsuite
MK: Also brings up John Snow's epidemiology study, closing the pump. Today 'We'd use GPS' #ASHG14
12:14pm October 19th 2014 via Hootsuite
MK:Mentions Ginsburg's work on H3N2 at Duke for infection models, figure from this paper http://t.co/l27yxiYc5Q #ASHG14
12:13pm October 19th 2014 via Hootsuite
MK:Mentions Watson in a recent Economist piece http://t.co/isj9be1ytC The 'exposome' of environmental variables #ASHG14
12:10pm October 19th 2014 via Hootsuite
MK: An IOM ecological model and need for multilevel analyses of 'causation'. The Genome is 'just the beginning' #ASHG14
12:08pm October 19th 2014 via Hootsuite
MK:As a text mining tool, points to 2010 paper http://t.co/yfVI6qn5HL #ASHG14
12:07pm October 19th 2014 via Hootsuite
MK @DrKhouryCDC: Big data: 'CDC HuGE Navigator': based upon PubMed, started in 2000. Website: http://t.co/vGgBMm1HN7 #ASHG14
12:05pm October 19th 2014 via Hootsuite
Last up: Muin Khoury (CDC Ofc of Public Health Genomics): Medicine and Public Health #ASHG14
12:03pm October 19th 2014 via Hootsuite
DG:Q:Back to N=1 Mass customization? A:N of many needs tools & standards as first step. As tools mature - to test & implement #ASHG1
12:02pm October 19th 2014 via Hootsuite
