Inferring ongoing cancer evolution from single tumour biopsies using synthetic supervised learning.

Variant allele frequencies (VAF) encode ongoing evolution and subclonal selection in growing tumours. However, existing methods that utilize VAF information for cancer evolutionary inference are compressive, slow, or incorrectly specify the underlying cancer evolutionary dynamics. Here, we provide a proof-of-principle synthetic supervised learning method, TumE, that integrates simulated models of cancer evolution with Bayesian neural networks, to infer ongoing selection in bulk-sequenced single tumour biopsies. Analyses in synthetic and patient tumours show that TumE significantly improves both accuracy and inference time per sample when detecting positive selection, deconvoluting selected subclonal populations, and estimating subclone frequency. Importantly, we show how transfer learning can leverage stored knowledge within TumE models for related evolutionary inference tasks-substantially reducing data and computational time for further model development and providing a library of recyclable deep learning models for the cancer evolution community. This extensible framework provides a foundation and future directions for harnessing progressive computational methods for the benefit of cancer genomics and, in turn, the cancer patient.

MaveQuest: a web resource for planning experimental tests of human variant effects.

SUMMARY: Fully realizing the promise of personalized medicine will require rapid and accurate classification of pathogenic human variation. Multiplexed assays of variant effect (MAVEs) can experimentally test nearly all possible variants in selected gene targets. Planning a MAVE study involves identifying target genes with clinical impact, and identifying scalable functional assays for that target. Here, we describe MaveQuest, a web-based resource enabling systematic variant effect mapping studies by identifying potential functional assays, disease phenotypes and clinical relevance for nearly all human protein-coding genes. AVAILABILITY AND IMPLEMENTATION: MaveQuest service: https://mavequest.varianteffect.org/. MaveQuest source code: https://github.com/kvnkuang/mavequest-front-end/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Integrating disease and drug-related phenotypes for improved identification of pharmacogenomic variants.

Aim: To improve the identification and interpretation of pharmacogenetic variants through the integration of disease and drug-related traits. Materials & methods: We hypothesized that integrating genome-wide disease and pharmacogenomic data may drive new insights into drug toxicity and response by identifying shared genetic architecture. Pleiotropic variants were identified using a methodological framework incorporating colocalization analysis. Results: Using genome-wide association studies summary statistics from the UK Biobank, European Bioinformatics Institute genome-wide association studies catalog and the Pharmacogenomics Research Network, we validated pleiotropy at the ABCG2 locus between allopurinol response and gout and identified novel pleiotropy between antihypertensive-induced new-onset diabetes, Crohn's disease and inflammatory bowel disease at the IL18RAP/SLC9A4 locus. Conclusion: New mechanistic insights and genetic loci can be uncovered by identifying pleiotropy between disease and drug-related traits.

Lay abstract Disease-focused genome-wide association studies (GWAS) have identified a plethora of actionable genetic variants over the last 20 years. International collaboration and technological breakthroughs have enabled rapid genotyping and data collection, which has correspondingly increased sample size and power. Contrastingly, recruitment of well-characterized cohorts of patients for pharmacogenomics research has proven challenging. Given the greater number of associated genetic variants and larger cohort sizes in common disease GWAS, we hypothesized that integrating genome-wide disease and pharmacogenomic data may drive new insights into drug toxicity and drug efficacy phenotypes, beyond the standard scope of current pharmacogenetic analyses. Using GWAS summary statistics from the UK Biobank, European Bioinformatics Institute GWAS catalog, and the Pharmacogenomics Research Network, and a methodological framework incorporating colocalization analysis, we validated pleiotropy at the ABCG2 locus between allopurinol response, gout, and serum urate and identified novel pleiotropy between antihypertensive-induced new-onset diabetes, Crohn's disease and inflammatory bowel disease at the IL18RAP/SLC9A4 locus.