flâneur — a map of the web's best reading

Our hetnet edge prediction methodology: the modeling framework for Project Rephetio | Thinklab

think-lab.github.io · 2,727 words · saved by 1 readers

This discussion will describe the machine learning pipeline used by Project Rephetio to predict the probability that a compound treats a disease. Our method derives from a 2011 study to predict coauthorship [1] and our 2015 study to predict disease–gene associations [2]. Several months of investigation by @alizee, @sergiobaranzini, and I have led to several advances in our algorithm, which we call metapath-based hetnet edge prediction or HNEP for short.

Our hetnet edge prediction methodology: the modeling framework for Project Rephetio | Thinklab --> --> Browse Proposals Projects Article discussion All discussion Leaderboard About Login Join --> --> Project: Rephetio: Repurposing drugs on a hetnet [rephetio] Our hetnet edge prediction methodology: the modeling framework for Project Rephetio Daniel Himmelstein Researcher May 4, 2016 Background This discussion will describe the machine learning pipeline used by Project Rephetio to predict the probability that a compound treats a disease. Our method derives from a 2011 study to predict coauthors

Explore this link on the map →

related reading