We provide two set of intronic and intergenic TAGOOS scores. The TAGOOS underlying model has been trained with the XGBOOST algorithm using intronic and intergenic associated SNPs (GWAS P-value < 0.8) from the GRASP database. The predictive variables have been selected by the learning algorithm among 4684 gene regulation related annotations such as histone modifications, eQTLs or transcription factors in different tissues from the Expressed enhancers, ENCODE, GTEx, ReMap, RoadMap and H3K27ac data sets.
The TAGOOS score documentation with user guide and API reference is hosted at Read The Docs.
To build the documentation, it is better to create a conda environment:
conda create --name tagoos python=3.5
pip install sphinx sphinx_rtd_theme recommonmarkThen run make html in the docs folder