Gillman, R. | bioRxiv Science

bioRxiv2026

Advancing Liver Cancer Precision Medicine with TARGET-SL

Background and AimsA major goal of personalised liver oncology is the ability to make targeted predictions about cancer-specific toxicity, however there are limited methods available. To address this, we validated the performance of our bioinformatics framework, TARGET-SL, through ex vivo drug screening. MethodsUsing TARGET-SL we predicted gain of function (GOF), loss of function (LOF) and synthetic lethal (SL) genetic events, and corresponding drug candidates. We validated drug predictions across hepatocellular carcinoma (HCC) cell lines, and a cohort of HCC and cholangiocarcinoma (CCA) patient-derived organoids (PDOs). ResultsFor HCC cells and PDOs we found 37.5% and 25% of the respective selected compounds induced unique target-specific growth inhibition based on genetic biomarkers, suggesting novel biomarker-driven drug sensitivities. ConclusionsOur analyses demonstrate TARGET-SLs potential to enhance personalized drug screening for liver cancer, by focusing on genetically informed targets. This will reduce experimental costs and accelerate the pace of therapeutic discovery. Impact and ImplicationsPrimary liver cancer (PLC) is a cancer with poor prognosis, and current therapies increase survival only for a minority of patients. Through the application of TARGET-SL we can predict, for each patient, the essential genes and corresponding small molecule inhibitors. These data support further investigation in larger patient cohorts and offer the possibility to specify new small molecule inhibitors and to repurpose current drugs for PLC treatment. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=81 SRC="FIGDIR/small/725819v1_ufig1.gif" ALT="Figure 1"> View larger version (28K): org.highwire.dtl.DTLVardef@10cb252org.highwire.dtl.DTLVardef@8f3049org.highwire.dtl.DTLVardef@ab4467org.highwire.dtl.DTLVardef@17f9d3_HPS_FORMAT_FIGEXP M_FIG C_FIG HighlightsO_LITARGET-SL can predict gene and drug sensitivities for cell lines and patient-derived organoids C_LIO_LIThis may reduce drug screening costs and accelerate the pace of therapeutic discovery. C_LIO_LITARGET-SL may assist in the repurposing of current drugs and their rapid translation for primary liver cancer C_LIO_LITARGET-SL is tumour-type agnostic, and therefore may have application in other cancers with poor prognosis C_LI

cancer biology↗

bioRxiv2025

Detecting Somatic Mutations in Rare Clones using Single Cell Multi-Omics

BackgroundSomatic mutations are increasingly recognised as drivers of diseases beyond cancer, including autoimmune disorders. However, identifying rare, cell type-specific causal mutations remains challenging due to their low frequency within heterogeneous cell populations. Traditional bulk sequencing approaches lack the resolution to detect rare variants, underscoring the need for novel methods specifically designed for single-cell data of heterogeneous cell populations. MethodsWe present an integrated single-cell multi-omics computational framework, SCARCE (Single-Cell Analysis of Rare Clonal Events), tailored for single-cell DNA sequencing (scDNA-seq) to statistically prioritise rare somatic mutations within defined cell subpopulations. By comparing variant frequencies across subpopulations, identified through either variant-based clustering or cell type annotation from surface marker expression, we identify variants enriched in specific cell populations. Our method applies multiple user-adjustable filters and statistical enrichment tests to distinguish true somatic variants from technical artifacts. ResultsSCARCE successfully identifies rare somatic mutations across three distinct datasets using technologies including MissionBio Tapestri and clonally-amplified whole-genome sequencing of single cells. We demonstrate that SCARCE correctly isolates and identifies true variants in a cell population comprising just 10 of 16,316 cells (0.06% of the total population). Furthermore, in an extensively characterised sample with known causal variants, SCARCE correctly identifies all known pathogenic variants among its top-ranked candidates. ConclusionsSCARCE offers several advantages over existing tools in the field. By integrating genetic and phenotypic information at single-cell resolution, our approach opens new avenues for understanding the clonal origins of diseases driven by somatic mutations in small cell populations.

bioinformatics↗