Kaeli A. Rizzo

Abstract Deciphering how DNA sequence encodes gene regulation remains a central challenge in biology. Advances in machine learning and functional genomics have enabled sequence-to-function (seq2func) models that predict molecular regulatory readouts directly from DNA sequence, supporting variant effect prediction, mechanistic interpretation and regulatory sequence design.

Abstract Deep neural networks (DNNs) have advanced predictive modeling for regulatory genomics, but challenges remain in ensuring the reliability of their predictions and understanding the key factors behind their decision making. Here we introduce DEGU (Distilling Ensembles for Genomic Uncertainty-aware models), a method that integrates ensemble learning and knowledge distillation to improve the robustness and explainability of DNN predictions.

Predicting gene associations to MGI mouse phenotypes The MGI Phenotype Database is a comprehensive resource for characterizing single gene knockout phenotypes in mice (Law & Shaw, 2018). PrismEXP can predict mouse phenotypes for genes with a high level of accuracy, specifically, with an average AUC of 0.82. To demonstrate this, we visualize the feature space of the top predicted genes for the MGI Mouse Phenotype MP:0005134 Decreased Thyroid-Stimulating Hormone Level (Fig. 5).