)
)
Explainable Artificial Intelligence Identifies Novel Regulatory Regions in the Human Genome and Their Epigenetic Code
-
Majority of mutations associated with complex diseases (e.g., cancer) are often located in enhancers, non-coding regulatory elements in the genome that control gene expression.
-
Our AI models classified novel enhancers, and we have further experimentally validated a subset of these enhancers using both global epigenetic perturbations and directed enhancer epigenetic rewriting.
-
We generated rule-based eXplainable Artificial Intelligence (XAI) and several Machine Learning methods (including Deep Learning and XGBoost) for the enhancer prediction in several human and mouse cell lines which display high accuracy.
