CaClust: linking genotype to transcriptional heterogeneity of follicular lymphoma using BCR and exomic variants

Tumor tissues exhibit high genotypic and transcriptional heterogeneity, resulting from tumor evolution and affecting cancer progression and treatment. These two types of heterogeneity in follicular lymphoma were so far predominantly studied in separation. To comprehensively investigate the evolution and genotype to phenotype maps in follicular lymphoma, we introduce CaClust, a probabilistic graphical model that integrates deep whole exome, single-cell RNA and B-cell receptor sequencing data to infer clone genotypes, cell-to-clone mapping, and single-cell genotyping. CaClust outperforms a state-of-the-art model on simulated and patient data. In-depth analysis of 22492 single cells and whole exomes from four follicular lymphoma samples using CaClust gives insights into effects of driver mutations, follicular lymphoma evolution, and possible therapeutic targets. CaClust single-cell genotyping agrees with genotypes observed in an independent targeted resequencing experiment. Our approach is the first to evaluate the strength of genotype to phenotype links in follicular lymphoma in the evolutionary context of the disease.