Background: Breast cancer is intrinsically heterogeneous and is commonly classified into four main subtypes associated with distinct biological features and clinical outcomes. However, currently available data resources and methods are limited in identifying molecular subtyping on protein-coding genes, and little is known about the roles of long non-coding RNAs (lncRNAs), which occupies 98% of the whole genome. lncRNAs may also play important roles in subgrouping cancer patients and are associated with clinical phenotypes. Methods: The purpose of this project was to identify lncRNA gene signatures that are associated with breast cancer subtypes and clinical outcomes. We identified lncRNA gene signatures from The Cancer Genome Atlas (TCGA )RNAseq data that are associated with breast cancer subtypes by an optimized 1-Norm SVM feature selection algorithm. We evaluated the prognostic performance of these gene signatures with a semi-supervised principal component (superPC) method. Results: Although lncRNAs can independently predict breast cancer subtypes with satisfactory accuracy, a combined gene signature including both coding and non-coding genes will give the best clinically relevant prediction performance. We highlighted eight potential biomarkers (three from coding genes and five from non-coding genes) that are significantly associated with survival outcomes. Conclusion: Our proposed methods are a novel means of identifying subtype-specific coding and non-coding potential biomarkers that are both clinically relevant and biologically significant.

A number of recent studies have highlighted the findings that certain lncRNAs are associated with alternative splicing (AS) in tumorigenesis and progression. Although existing work showed the importance of linking certain misregulations of RNA splicing with lncRNAs, a primary concern is the lack of genome-wide comprehensive analysis for their associations.We analyzed an extensive collection of RNA-seq data, quantified 198,619 isoform expressions, and found systematic isoform usage changes between hepatocellular carcinoma (HCC) and normal liver tissue. We identified a total of 1375 splicing switched isoforms and further analyzed their biological functions.To predict which lncRNAs are associated with these AS genes, we integrated the co-expression networks and epigenetic interaction networks collected from text mining and database searching, linking lncRNA modulators such as splicing factors, transcript factors, and miRNAs with their targeted AS genes in HCC. To model the heterogeneous networks in a single framework, we developed a multi-graphic random walk (RWMG) network method to prioritize the lncRNAs associated with AS in HCC. RWMG showed a good performace evaluated by ROC curve based on cross-validation and bootstrapping strategy.As a summary, we identified 31 AS-related lncRNAs including MALAT1 and HOXA11-AS, which have been reported before, as well as some novel lncRNAs such as DNM1P35, HAND2-AS1, and DLX6-AS1. Survival analysis further confirmed the clinical significance of identified lncRNAs.