Scientist. He/Him.
Derwood, MD
Joined March 2011
- Tweets91
- Following847
- Followers66
- Likes233
Pinned Tweet
Our paper about SARS-CoV-2 genomic determinants of pathogenicity is out.
Read our paper here: pnas.org/content/early/2020/…
NIH Press release: nih.gov/news-events/news-rel…
Our paper using #MachineLearning for prediction of virus incubation periods (including #SARSCoV2) based solely on their genomes is out! We used an elastic net regression model (via @scikit_learn) to predict the incubation periods, with a mean error of +/- 1.63 days.
New #openaccess research from @eugene_koonin and colleagues: Prediction of the #incubation period for #COVID19 and future virus disease outbreaks. #MachineLearning
Read it here: bmcbiol.biomedcentral.com/ar…
New video from @IRPatNIH in which I describe our exploration of the genomic characteristics that may make SARS-CoV-2 deadly #COVID19 #coronavirus irp.nih.gov/scibites/identif…
Our paper leveraging machine-learning for the discovery of novel anti-CRISPR proteins is out! With @joeBondyDenomy, @eugene_koonin and others
nature.com/articles/s41467-0…
Congrats @dimvitsios! Really exciting work!
Really excited to see our latest work out at The American Journal of Human Genetics!🙂@AJHGNews
cell.com/ajhg/fulltext/S0002…
mantis-ml is an auto-ML framework for gene prioritisation across any disease of interest.
GitHub: github.com/astrazeneca-cgr-p…
#MachineLearning @AstraZeneca
1
1
Our preprint about SARS-CoV-2 genomic determinants of pathogenicity is live (biorxiv.org/content/10.1101/…). We identified key genomic features associated with the high fatality rates and the host switch from animals to humans in coronaviruses.
Check out our new preprint where we use machine learning to discover new anti-CRISPR proteins! biorxiv.org/content/10.1101/…
Of 38 methods examined in this paper, our method for predicting the pathogenicity of noncoding variants (Orion bit.ly/33XAWPg) was shown to be the very top feature! (go.nature.com/2E4usnj, Fig. 1C)
hey @ryandhindsa @medchances should we add "willing to bribe" to our medical school admissions predictors? may be a strong feature for medchances.org #CollegeCheatingScandal #collegeadmissionsscandal
3 steps to building the perfect medical school application list medchances.org/blog/3-steps-…
1
@camphalfblood May Orion, our algorithm that hunts for disease in genome dark regions, hunt as well as its namesake!
newsroom.cumc.columbia.edu/b…
1
We explore the genome's dark regions at @DukeU/@Columbia in @PLOSONE study journals.plos.org/plosone/ar… and @DailyMailUK dailymail.co.uk/health/artic…
2