Framework for Interpretable Neural Networks for Genetics GitHub

In Just DNA-Seq analysis pipeline for personal and public genomes, we use trained models from GenNet paper for diseases that can be associated with aging like diabetes, Atrial fibrillation (AF), hypertension, breast cancer, dementia, and balding. The models were trained on UK Biobank data using 6,986,636 input variants from whole exome sequencing (WES) data.

Both 'gennet_launch_.py' and 'gennet_launch_.ipynb' can be used to launch a model. For various types of architecture, files are complemented with appropriate masks. Python 3.7 and TensorFlow 2.8.0 are needed, see full instructions on the GenNet GitHub page.