Preprint introduces TransCisPredict tool for biobank-scale proteome-wide association studies

Proteome-wide association studies (PWAS) aim to connect variation in protein abundance to disease risk, but most large-scale genetic studies lack direct proteomic measurements. A preprint posted to bioRxiv describes TransCisPredict, a new computational framework designed to address this gap by predicting protein expression levels from genetic data and enabling PWAS at biobank scale.

The method reduces computational burden through linkage-disequilibrium block selection and incorporates both cis-acting variants — those near the gene encoding the protein — and trans-acting variants, which may exert effects from elsewhere in the genome. The authors report that accounting for trans-variants improves prediction accuracy for a subset of proteins with complex regulatory architectures, and validate the approach by performing protein-phenotype association analyses in a large biobank dataset.

TransCisPredict is likely to be of interest to researchers working in statistical genetics, functional genomics, and the expanding field of multi-omics integration. PWAS complements transcriptome-wide association studies (TWAS) and is seen as one route to prioritising causal genes at GWAS loci by moving closer to the molecular effectors of disease.

The preprint has not yet undergone peer review. Code availability details are described in the preprint.