A Type-2 Fuzzy Logic-Based Explainable Artificial Intelligence for the Prediction of Enhancers

Enhancers are the non-coding regions of the genome responsible for controlling the activity of the genes. Approximately 80-90% of mutations causing human diseases (including cancers) are located in the non-coding part of the human genome, often in enhancers. To provide a better understanding of how these mutations lead to disease states, machine and deep learning approaches have been developed to annotate enhancers. However, these opaque box models lack transparency and fail to provide genomics and clinical researchers with an accurate explanation of how these models predict specific regions as enhancers over other regions. Hence, there is a need for eXplainable Artificial Intelligence (XAI) systems that can be easily understood, analysed, and augmented by domain experts. In this paper, we will present a type-2 fuzzy logic-based XAI system to predict the enhancer regions which was trained and applied in the Drosophila genome. The model provides explainable rules that are easily understandable. The model has significantly recovered the experimentally validated enhancers, along with the large set of novel enhancers that display similar characteristics to the experimentally validated enhancers. This research is the first step to provide a mechanistic understanding at a molecular level of how the genome is regulated.