Here we show that TATES has correct false positive (type-I error) rate, and that TATES picks up both phenotype-specific genetic effects as well as genetic effects that are common to multiple correlated phenotypes. In particular, this facili-tates representing the NCA in standard Tensorflow code. In principle, kernels can be learnable rather than hard-coded; however this makes the trained models less interpretable and so we do not pursue this approach here.

Understanding the Context

The specific risk of sexual violence and exploitation among WMSSA creates a substantial barrier for providers in effectively discussing PrEP, as it necessi-tates addressing a sensitive issue that can foster a negative mindset around PrEP use in women. We developed a Random-Forest based algorithm, RFECS (R andom F orest based E nhancer identification from C hromatin S tates) to integrate histone modification profiles for identification of enhancers, and used it to identify enhancers in a number of cell-types.