ResidualBind: Uncovering Sequence-Structure Preferences of RNA-Binding Proteins with Deep Neural Networks.

Deep neural networks have demonstrated improved performance at predicting sequence specificities of DNA- and RNA-binding proteins. However, it remains unclear why they perform better than previous methods that rely on k-mers and position weight matrices. Here, we highlight a recent deep learning-based software package, called ResidualBind, that analyzes RNA-protein interactions using only RNA sequence as an input feature and performs global importance analysis for model interpretability. We discuss practical considerations for model interpretability to uncover learned sequence motifs and their secondary structure preferences.

Reassessing pharmacogenomic cell sensitivity with multilevel statistical models.

Pharmacogenomic experiments allow for the systematic testing of drugs, at varying dosage concentrations, to study how genomic markers correlate with cell sensitivity to treatment. The first step in the analysis is to quantify the response of cell lines to variable dosage concentrations of the drugs being tested. The signal to noise in these measurements can be low due to biological and experimental variability. However, the increasing availability of pharmacogenomic studies provides replicated data sets that can be leveraged to gain power. To do this, we formulate a hierarchical mixture model to estimate the drug-specific mixture distributions for estimating cell sensitivity and for assessing drug effect type as either broad or targeted effect. We use this formulation to propose a unified approach that can yield posterior probability of a cell being susceptible to a drug conditional on being a targeted effect or relative effect sizes conditioned on the cell being broad. We demonstrate the usefulness of our approach via case studies. First, we assess pairwise agreements for cell lines/drugs within the intersection of two data sets and confirm the moderate pairwise agreement between many publicly available pharmacogenomic data sets. We then present an analysis that identifies sensitivity to the drug crizotinib for cells harboring EML4-ALK or NPM1-ALK gene fusions, as well as significantly down-regulated cell-matrix pathways associated with crizotinib sensitivity.

Improving representations of genomic sequence motifs in convolutional networks with exponential activations.

Deep convolutional neural networks (CNNs) trained on regulatory genomic sequences tend to build representations in a distributed manner, making it a challenge to extract learned features that are biologically meaningful, such as sequence motifs. Here we perform a comprehensive analysis on synthetic sequences to investigate the role that CNN activations have on model interpretability. We show that employing an exponential activation to first layer filters consistently leads to interpretable and robust representations of motifs compared to other commonly used activations. Strikingly, we demonstrate that CNNs with better test performance do not necessarily imply more interpretable representations with attribution methods. We find that CNNs with exponential activations significantly improve the efficacy of recovering biologically meaningful representations with attribution methods. We demonstrate these results generalise to real DNA sequences across several in vivo datasets. Together, this work demonstrates how a small modification to existing CNNs, i.e. setting exponential activations in the first layer, can significantly improve the robustness and interpretabilty of learned representations directly in convolutional filters and indirectly with attribution methods.

Deep learning for inferring transcription factor binding sites.

Deep learning is a powerful tool for predicting transcription factor binding sites from DNA sequence. Despite their high predictive accuracy, there are no guarantees that a high-performing deep learning model will learn causal sequence-function relationships. Thus a move beyond performance comparisons on benchmark datasets is needed. Interpreting model predictions is a powerful approach to identify which features drive performance gains and ideally provide insight into the underlying biological mechanisms. Here we highlight timely advances in deep learning for genomics, with a focus on inferring transcription factors binding sites. We describe recent applications, model architectures, and advances in local and global model interpretability methods, then conclude with a discussion on future research directions.