FABIAN-variant 2026: improved prediction of the effects of DNA variants on transcription factor binding

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: Approximately 8-fold increase in human TF model coverage; addition of 1,290 TFFMs and predictions for over 1,400 heterodimer TF complexes; BPNet deep learning scorer added for 240 human TFs

• Primary Outcome: Expansion of TF binding prediction model library from approximately 5,000 to over 40,000 models covering more than 1,500 human TFs, with addition of mouse genome support (>35,000 models for >1,100 mouse TFs)

• Nnt Or Sensitivity: Not applicable in traditional clinical sense; runtime reduction of approximately 70% reported as a performance metric; no sensitivity/specificity or AUC values reported in the abstract

• Confidence Interval: Not reported

Clinical Application

High feasibility for integration into bioinformatics pipelines: free web access, RESTful JSON API, and standalone command-line version support both individual variant queries and high-throughput batch analysis. No computational infrastructure requirements for web-based use. Requires bioinformatics expertise to interpret outputs appropriately and integrate with clinical variant classification frameworks (e.g., ACMG/AMP criteria). Relevant to Australian clinical genomics practice, particularly within the context of the Australian Genomics Health Alliance, Mackenzie's Mission reproductive carrier screening programme, and Medicare-funded genomic testing pathways. Non-coding variant interpretation is an emerging priority for Australian clinical genetics services as whole-genome sequencing adoption increases. FABIAN-variant 2026 could support annotation of regulatory variants in conditions where coding variant screening is unrevealing. The tool is not a TGA-regulated medical device and does not require PBS listing; it functions as a research-grade in silico annotation tool. RACGP and HGSA guidelines emphasise that in silico predictions alone are insufficient for clinical variant classification and must be integrated with functional evidence — a caveat that applies directly to this tool's outputs. Clinical and research genomics laboratories interpreting non-coding variants (promoter and enhancer region variants) in patients undergoing diagnostic genome or whole-genome sequencing, particularly where variants of uncertain significance (VUS) in regulatory regions require functional annotation

Abstract

Variants in promoters and enhancers can alter the binding of transcription factors (TFs), but their functional assessment remains difficult. FABIAN-variant is a web application that predicts the effects of DNA variants on TF binding by comparing position weight matrix (PWM) and transcription factor flexible model (TFFM) scores between reference and variant alleles. Here, we present FABIAN-variant 2026, a major update that expands the prediction model library from ~5000 to over 40 000 models for >1500 human TFs, sourced from nine PWM databases and including 1290 TFFMs. The application now supports the mouse genome (GRCm38 and GRCm39) with over 35 000 models for >1100 mouse TFs. An optional BPNet deep learning scorer provides neural network-based binding predictions for 240 human TFs. Known TF binding site information has been expanded from three to five sources. Predictions for over 1400 heterodimer TF complexes have been added. The web server has been rewritten in Rust and the scoring engine optimized, reducing runtime by ~70%. A RESTful JSON API and a standalone command-line version enable programmatic access and local high-throughput analysis. FABIAN-variant 2026 is available at https://fabianapp.org/variant26/. The web server is free and open to all users and there is no login requirement.

References

1. 1.Steinhaus, R., Robinson, P. N., & Seelow, D. (2026). FABIAN-variant 2026: improved prediction of the effects of DNA variants on transcription factor binding. Nucleic Acids Research. PubMed ID: 42411323. Retrieved from https://pubmed.ncbi.nlm.nih.gov/42411323