Interpretable machine learning for genomics

Abstract

High-throughput technologies such as next-generation sequencing allow biologists to observe cell function with unprecedented resolution, but the resulting datasets are too large and complicated for humans to understand without the aid of advanced statistical methods. Machine learning (ML) algorithms, which are designed to automatically find patterns in data, are well suited to this task. Yet these models are often so complex as to be opaque, leaving researchers with few clues about underlying mechanisms. Interpretable machine learning (iML) is a burgeoning subdiscipline of computational statistics devoted to making the predictions of ML models more intelligible to end users. This article is a gentle and critical introduction to iML, with an emphasis on genomic applications. I define relevant concepts, motivate leading methodologies, and provide a simple typology of existing approaches. I survey recent examples of iML in genomics, demonstrating how such techniques are increasingly integrated into research workflows. I argue that iML solutions are required to realize the promise of precision medicine. However, several open challenges remain. I examine the limitations of current state-of-the-art tools and propose a number of directions for future research. While the horizon for iML in genomics is wide and bright, continued progress requires close collaboration across disciplines.

Mapping Data to Deep Understanding: Making the Most of the Deluge of SARS-CoV-2 Genome Sequences

“…Interpretable, or explainable, machine learning can fill this important gap ( 75 , 76 ). Interpretable machine learning is particularly important in bioinformatics, since explaining a model’s predictions is critical to justify making high-stakes clinical or research decisions based on machine learning predictions ( 77 , 78 ). Accordingly, developers of deep learning approaches to SARS-CoV-2 should consider providing some functionality to interpret or explain predictions.…”

An Interpretable Deep Learning Model for Predicting the Risk of Severe COVID-19 from Spike Protein Sequence

“…While neural networks excel at fitting data, they suffer from a "black box" problem: it is very hard to explain a neural network's predictions. 25 We add two layers to not only help classify data but also allow us to visualize and potentially interpret the models that we train. The first is a forward attention layer 26 , based on a structure used to analyze text 27 that can also be applied to biological sequences.…”

Access to online learning: Machine learning analysis from a social justice perspective

“…With the complexity of machine learning algorithms, interpretability does not come directly from the algorithm itself, but instead through an explanation model. For this, Shapley values have been proposed (Lundberg & Lee, 2017) which are 'model agnostic': that is, applicable to machine learning models, regardless of the algorithm used (Watson, 2021). Shapley values are fully 'additive' too, meaning that they possess all the properties relevant to additive feature attribution, and so completely "attribute[s] an effect to each feature and, [by] summing the effects of all feature attributions, approximates the output of the original [analytic] model" (Lundberg & Lee, 2017, p. 2).…”