Discovering novel mutation signatures by latent Dirichlet allocation with variational Bayes inference

Abstract

MotivationA cancer genome includes many mutations derived from various mutagens and mutational processes, leading to specific mutation patterns. It is known that each mutational process leads to characteristic mutations, and when a mutational process has preferences for mutations, this situation is called a ‘mutation signature.’ Identification of mutation signatures is an important task for elucidation of carcinogenic mechanisms. In previous studies, analyses with statistical approaches (e.g. non-negative matrix factorization and latent Dirichlet allocation) revealed a number of mutation signatures. Nonetheless, strictly speaking, these existing approaches employ an ad hoc method or incorrect approximation to estimate the number of mutation signatures, and the whole picture of mutation signatures is unclear.ResultsIn this study, we present a novel method for estimating the number of mutation signatures—latent Dirichlet allocation with variational Bayes inference (VB-LDA)—where variational lower bounds are utilized for finding a plausible number of mutation patterns. In addition, we performed cluster analyses for estimated mutation signatures to extract novel mutation signatures that appear in multiple primary lesions. In a simulation with artificial data, we confirmed that our method estimated the correct number of mutation signatures. Furthermore, applying our method in combination with clustering procedures for real mutation data revealed many interesting mutation signatures that have not been previously reported.Availability and implementationAll the predicted mutation signatures with clustering results are freely available at http://www.f.waseda.jp/mhamada/MS/index.html. All the C++ source code and python scripts utilized in this study can be downloaded on the Internet (https://github.com/qkirikigaku/MS_LDA).Supplementary information Supplementary data are available at Bioinformatics online.

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

“…The original data contains 2780 samples for single nucleotide substitution (SBS) counts. We removed samples with few mutations (less than 400 mutations) as a preprocessing because the presence of samples with low-variant interferes with learning in LDA-based approaches [ 16 ], and we obtained 2653 samples. Figure 3 shows the number of samples contained in each tumor type.…”

“…Figure 2 shows the graphical model of PLDA. A detailed notation of each parameter is provided in our previous study [ 16 ]. The vector is the parameter of K -dimensional multinomial distribution, indicating the signature activity of the s th sample in l th tumour type ( ).…”

“…The variational Bayes method updates the parameters ( in Figure 2 ) to improve the objective function, variational lower bound, which is also utilized during model selection (cf. see [ 16 ]). The bottleneck part of the computational complexity with respect to time is the calculation of the responsibility from which signature each mutation is generated, , and it follows .…”

“…The signature activity of l th tumor type, , is generated from Dirichlet distribution with . Other parameters other than represent the same object as the method using normal LDA [ 16 ].…”

“…In this matrix decomposition, it is necessary to regularize for likelihood and select the number of signatures ( K in Figure 1 ) contributing to the composition of mutation catalogs because the degree of freedom of the model and the likelihood after parameter estimation continuously increases with an increase in K . To resolve this issue, EMu [ 13 ] and signeR [ 14 ] using BIC, which is one of the information criteria, and the method using latent Dirichlet allocation [ 15 , 16 ] have been developed, enabling prediction with greater mathematical stringency. Some mutation signatures are being confirmed through biological assays and are no longer simply concepts derived from in silico analyses, e.g., Zou et al (2018) used CRISPR-Cas9 technology to knock out individual genes in vitro, and confirmed the existence of mutation signatures [ 17 ].…”