TEffectBayes: A Nextflow Pipeline for Exploring the Potential Effect of Transposable Elements in Gene Regulatory Network with Multi-Omic Bayesian Network Model

Abstract

Transposable elements (TEs) are critical contributors to gene regulatory networks, yet their repetitive and abundant nature complicates efforts to elucidate their precise regulatory roles. While existing computational tools facilitate systematic identification of associations between TEs and gene expression, these methods typically cannot account for confounding variables or capture causal and directional interactions. To address these limitations, we developed TEffectBayes, a Nextflow-based pipeline leveraging a multi-omic Bayesian network (BN) framework designed to systematically infer directional, probabilistic regulatory dependencies involving TEs. TEffectBayes integrates diverse omics datasets, including RNA-seq-derived gene and locus-specific TE expression, along with ChIP-seq-based histone modification data processed via custom R and Python scripts. Integrated multi-omic datasets are subsequently employed to build gene-centric Bayesian models, enabling robust inference of context-dependent, probabilistic relationships between TEs, chromatin modifications, and gene expression. TEffectBayes thus provides a reproducible and scalable computational framework for unraveling the complex regulatory landscape shaped by TEs. In summary, TEffectBayes supports systematic prioritization of TE-chromatin-gene regulatory candidates for downstream benchmarking and experimental validation, enabling hypothesis-driven follow-up studies in diverse biological contexts. The pipeline, along with comprehensive user tutorials and example datasets, is publicly accessible at https://github.com/nkaan-kutlu/TEffectBayes.