Causal network inference from gene transcriptional time-series response to glucocorticoids

Gene regulatory network inference is essential to uncover complex relationships among gene pathways and inform downstream experiments, ultimately enabling regulatory network re-engineering. Network inference from transcriptional time-series data requires accurate, interpretable, and efficient determination of causal relationships among thousands of genes. Here, we develop Bootstrap Elastic net regression from Time Series (BETS), a statistical framework based on Granger causality for the recovery of a directed gene network from transcriptional time-series data. BETS uses elastic net regression and stability selection from bootstrapped samples to infer causal relationships among genes. BETS is highly parallelized, enabling efficient analysis of large transcriptional data sets. We show competitive accuracy on a community benchmark, the DREAM4 100-gene network inference challenge, where BETS is one of the fastest among methods of similar performance and additionally infers whether causal effects are activating or inhibitory. We apply BETS to transcriptional time-series data of differentially-expressed genes from A549 cells exposed to glucocorticoids over a period of 12 hours. We identify a network of 2768 genes and 31, 945 directed edges (FDR <= 0.2). We validate inferred causal network edges using two external data sources: Overexpression experiments on the same glucocorticoid system, and genetic variants associated with inferred edges in primary lung tissue in the Genotype-Tissue Expression (GTEx) v6 project. BETS is available as an open source software package at . Author summary We can better understand human health and disease by studying the state of cells and how environmental dysregulation affects cell state. Cellular assays, when collected across time, can show us how genes in cells respond to stimuli. These time-series assays provide an opportunity to identify causal relationships among thousands of genes without performing hundreds of thousands of experiments. However, inferring causal relationships from these time-series data needs to be fast, robust, and accurate. We present a method, BETS, that infers causal gene networks from gene expression time series. BETS runs quickly because it is parallelized, allowing even data sets with thousands of genes to be analyzed. We demonstrate the performance of BETS compared to 21 other state-of-the-art inference methods on benchmark data. We then use BETS to build causal networks from gene expression responses to the widely-prescribed drug dexamethasone. We replicate the estimated causal relationships using gene expression data from the Genotype-Tissue Expression (GTEx) project and from additional experiments with dexamethasone. We release our software so that BETS can be used to accurately and effectively infer causal relationships from gene expression time-series assays.

Automated summary

BETS is a method for inferring causal gene networks from gene expression time series, its efficiency and parallelization allow for quick analysis of large datasets and competitive performance in benchmark testing. Through external data validation, BETS can accurately infer activating or inhibitory causal effects.