DOI: https://doi.org/10.21203/rs.2.19530/v1
Background: Cancer is a complex disease with a high rate of mortality. The characteristics of tumor masses are very heterogeneous; thus, the appropriate classification of tumors is a critical point in the correct treatment. A high level of heterogeneity has also been observed in breast cancer. Therefore, detecting the molecular subtypes of this disease is a worthwhile issue for medicine that could be facilitated using bioinformatics.
Method: Numerous methods have already classified breast cancer based on gene expression data; however, they are not reliable due to the dynamic nature of these data. In contrast, gene mutation data are relatively stable and may lead to better classification. The aim of this study is to introduce a novel method for detecting the molecular subtypes of breast cancer. In this study, somatic mutation profiles of tumors are used; nonetheless, the somatic mutation profiles are very sparse. To address this issue, we made use of the network propagation method on gene interaction network and made the mutation profiles dense. Afterward, we used deep embedded clustering (DEC) method to classify breast tumors into four subtypes. In the next step, gene signatures of each subtype obtained by Fisher exact test and Benjamini-Hochberg procedure.
Results: Clinical and molecular analyses are executed, besides enrichment of results in numerous databases have shown that the proposed method, using mutation profiles can efficiently detect the molecular subtypes of breast cancer. Finally, a supervised classifier is proposed based on discovered subtypes to predict the molecular subtype of a new patient.
This preprint is available for download as a PDF.