Parallelized Latent Dirichlet Allocation Provides a Novel Interpretability of Mutation Signatures in Cancer Genomes

Abstract

Mutation signatures are defined as the distribution of specific mutations such as activity of AID/APOBEC family proteins. Previous studies have reported numerous signatures, using matrix factorization methods for mutation catalogs. Different mutation signatures are active in different tumor types; hence, signature activity varies greatly among tumor types and becomes sparse. Because of this, many previous methods require dividing mutation catalogs for each tumor type. Here, we propose parallelized latent Dirichlet allocation (PLDA), a novel Bayesian model to simultaneously predict mutation signatures with all mutation catalogs. PLDA is an extended model of latent Dirichlet allocation (LDA), which is one of the methods used for signature prediction. It has parallelized hyperparameters of Dirichlet distributions for LDA, and they represent the sparsity of signature activities for each tumor type, thus facilitating simultaneous analyses. First, we conducted a simulation experiment to compare PLDA with previous methods (including SigProfiler and SignatureAnalyzer) using artificial data and confirmed that PLDA could predict signature structures as accurately as previous methods without searching for the optimal hyperparameters. Next, we applied PLDA to PCAWG (Pan-Cancer Analysis of Whole Genomes) mutation catalogs and obtained a signature set different from the one predicted by SigProfiler. Further, we have shown that the mutation spectrum represented by the predicted signature with PLDA provides a novel interpretability through post-analyses.

Global Mapping of Interventions to Improve Quality of Life of Patients with Cancer: A Protocol for Literature Mining and Meta-Analysis

“…In this study, the model of the Latent Dirichlet Allocation (LDA) command in STATA was applied. The LDA is a familiar and helpful method for topic classification modeling into a group of similar topics [ 38 ]. It can separate the collected scientific research as “discrete distributions over latent topics”; each topic can perform as a “discrete distribution over all the terms”.…”

Clone decomposition based on mutation signatures provides novel insights into mutational processes

“…Besides, SigTracer prepares Dirichlet distributions as conjugate priors for the signature activity of each clone. This extension can be regarded as a generalization of the CloneSig model, and other studies on similar tasks such as signature extraction have already highlighted the effectiveness of Dirichlet priors [14,15,16]. Here, we aimed to show if SigTracer provides more reasonable clone estimations than CloneSig for artificial tumors.…”