Predicting gene expression with Machine Learning: an explanation in simple terms

Some of the ML papers in the genomics space can at first seem intimidating, since there is a vast amount of domain knowledge necessary to fully understand them. Here I briefly introduce some of the challenges in interpreting the genome before describing one of the most recent advances in Machine Learning applied to the gene expression prediction problem. My focus is not as much in the model architecture (though it is covered at a high level), but in understanding why such a model is useful, the kinds of data that were used for training it, and the types of outputs it provides. Understanding the problems that have been solved with ML, those that still exist, as well as the practical applications of existing approaches is critical to understand how the work we do as ML practitioners helps move the field forward and accelerate our understanding of human evolution and disease. <a href="https://cgaitan.medium.com/predicting-gene-expression-with-machine-learning-an-explanation-in-simple-terms-c2f134b40598">Website</a>