AUTHOR=Jiang Limin , Xiao Yongkang , Ding Yijie , Tang Jijun , Guo Fei 

TITLE=Discovering Cancer Subtypes via an Accurate Fusion Strategy on Multiple Profile Data

JOURNAL=Frontiers in Genetics

VOLUME=Volume 10 - 2019

YEAR=2019

URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2019.00020

DOI=10.3389/fgene.2019.00020

ISSN=1664-8021

ABSTRACT=Discovering cancer subtypes is useful for guiding clinical treatment on multiple cancers. Progressive profile technologies for tissue have accumulated diverse types of data. Based on these types of expression data, various computational methods have been proposed to predict cancer subtypes. It is inevitable to study how to better integrate these multiple profile data. In this paper, we collect multiple profile data for five cancers on The Cancer Genome Atlas (TCGA). Then, we construct three similarity kernels for all patients of same cancer by gene expression, miRNA expression and isoform expression data. What's more, we propose an novel unsupervised multiple kernel fusion method, Similarity Kernel Fusion (SKF), in order to integrate three similarity kernels into one combinated kernel. Finally, we make use of spectral clustering on the integrated kernel to predict cancer subtypes. In the experimental results, P value of Cox regression model and survival curve analysis can be used to evaluation the performance of predicted subtypes on three datasets. Our kernel fusion method, SKF, has outstanding performance compared with single kernel and other multiple kernel fusion strategies. It demonstrates that our method can accurately identify more accurate subtypes on various kinds of cancers. Our cancer subtypes prediction method can identify essential genes and biomarkers for disease diagnosis and prognosis, and also discuss the possible side effects of therapies and treatment.