An integrated approach to identify bimodal genes associated with prognosis in câncer.

Clovis Ferreira dos Reis,Josivan Ribeiro Justino,Sandro Jose de Souza,Beatriz Stransky,Andre Luis Fonseca

doi:10.1590/1678-4685-gmb-2021-0109

Abstract

Bimodal gene expression (where a gene expression distribution has two maxima) is associated with phenotypic diversity in different biological systems. A critical issue, thus, is the integration of expression and phenotype data to identify genuine associations. Here, we developed tools that allow both: i) the identification of genes with bimodal gene expression and ii) their association with prognosis in cancer patients from The Cancer Genome Atlas (TCGA). Bimodality was observed for 554 genes in expression data from 25 tumor types. Furthermore, 96 of these genes presented different prognosis when patients belonging to the two expression peaks were compared. The software to execute the method and the corresponding documentation are available at the Data access section.

Highlights

Studies on gene expression and regulation have been directed towards a better understanding of a diverse range of biological processes, including initial differentiation in the embryonic stage and changes in health and disease that occur during life
To ensure the reliable selection of samples according to the selected bimodal genes, an analysis based on the Gaussian Mixture Models (GMM) probabilistic model was introduced in our computational protocol
GMM has been previously used in the analysis of gene expression data (Ficklin et al, 2017; Golumbeanu et al, 2019; Mirzal, 2020) but to our knowledge this is the first application of such a method for the identification of genes with bimodal expression patterns

Summary

Introduction

Studies on gene expression and regulation have been directed towards a better understanding of a diverse range of biological processes, including initial differentiation in the embryonic stage and changes in health and disease that occur during life. These patterns of gene expression have been extensively used to establish associations between phenotypes and genetic/epigenetic information (Boyle et al, 2017; Young et al, 2019). The challenges for such studies are significant, and the identification of expression signatures enriched with genuine phenotypic associations is welcome. For an extensive review of the different methods developed for detection of bimodality, please see Moody et al (2019)

Methods

Results

Conclusion