thomas_sandmannAmariuta et a use ChIP-seq data from a single transcription factor to identify regulatory elements. Then they train a glmnet model on these examples to predict more regulatory elements based on broader epigenetic marks. Thanks for the careful description of the methods! #Geneticst.co/4opo5sC2mT
soumya_bostonRegulatory elements are defined by identifying chromatin signatures of key driving transcription factors, e.g. T-bet. Uses almost 400 genome-wide annotations, mostly functional epigenomic features from public sources.