Explainable Algorithms

Convolutional filtering for mutation signatures

A novel convolutional filter-based approach for learning mutation signature representations.

Neural Decomposition

Developing interpretable deep neural network approaches for dimensionality reduction that embed functional ANOVA decompositions within variational autoencoders.


Unravelling the history of genomic instability through deep reinforcement learning.

The Automatic BioData Scientist

Developing automated learning frameworks for reproducible and transferable biological data analysis.