ASpediaFI: Functional Interaction Analysis of Alternative Splicing Events.

Alternative splicing (AS) regulates biological processes governing phenotypes and diseases. Differential AS (DAS) gene test methods have been developed to investigate important exonic expression from high-throughput datasets. However, the DAS events extracted using statistical tests are insufficient to delineate relevant biological processes. In this study, we developed a novel application, Alternative Splicing Encyclopedia: Functional Interaction (ASpediaFI), to systemically identify DAS events and co-regulated genes and pathways. ASpediaFI establishes a heterogeneous interaction network of genes and their feature nodes (i.e., AS events and pathways) connected by co-expression or pathway gene set knowledge. Next, ASpediaFI explores the interaction network using the random walk with restart algorithm and interrogates the proximity from a query gene set. Finally, ASpediaFI extracts significant AS events, genes, and pathways. To evaluate the performance of our method, we simulated RNA sequencing (RNA-seq) datasets to consider various conditions of sequencing depth and sample size. The performance was compared with that of other methods. Additionally, we analyzed three public datasets of cancer patients or cell lines to evaluate how well ASpediaFI detects biologically relevant candidates. ASpediaFI exhibits strong performance in both simulated and public datasets. Our integrative approach reveals that DAS events that recognize a global co-expression network and relevant pathways determine the functional importance of spliced genes in the subnetwork. ASpediaFI is publicly available at https://bioconductor.org/packages/ASpediaFI.

Genomic Characteristics of Triple-Negative Breast Cancer Nominate Molecular Subtypes That Predict Chemotherapy Response.

The heterogeneity of triple-negative breast cancer (TNBC) poses difficulties for suitable treatment and leads to poor outcome. This study aimed to define a consensus molecular subtype (CMS) of TNBC and thus elucidate genomic characteristics and relevant therapy. We integrated the expression profiles of 957 TNBC samples from published datasets. We identified genomic characteristics of subtype by exploring the pathway activity, microenvironment, and clinical relevance. In addition, drug response (DR) scores (n = 181) were computationally investigated using chemical perturbation gene signatures and validated in our own patient with TNBC (n = 38) who received chemotherapy and organoid biobank data (n = 64). Subsequently, cooperative functions with drugs were also explored. Finally, we classified TNBC into four CMSs: stem-like; mesenchymal-like; immunomodulatory; luminal-androgen receptor. CMSs also elucidated distinct tumor-associated microenvironment and pathway activities. Furthermore, we discovered metastasis-promoting genes, such as secreted phosphoprotein 1 by comparing with primary. Computational DR scores associated with CMS revealed drug candidates (n = 18), and it was successfully evaluated in cisplatin response of both patients and organoids. Our CMS recapitulated in-depth functional and cellular heterogeneity encompassing primary and metastatic TNBC. We suggest DR scores to predict CMS-specific DRs and to be successfully validated. Finally, our approach systemically proposes a relevant therapeutic prediction model as well as prognostic markers for TNBC. IMPLICATIONS: We delineated the genomic characteristic and computational DR prediction for TNBC CMS from gene expression profile. Our systematic approach provides diagnostic markers for subtype and metastasis verified by machine-learning and novel therapeutic candidates for patients with TNBC.

Reproduction of molecular subtypes of gastric adenocarcinoma by transcriptome sequencing of archival tissue.

Gastric cancer (GC) is a heterogeneous disease, so molecular classification is important for selecting the most appropriate treatment strategies for GC patients. To be applicable in the clinic, there is an urgent need for a platform that will allow screening real-life archival tissue specimens. For this purpose, we performed RNA sequencing of 50 samples from our Asian Cancer Research Group (ACRG) GC cohort to reproduce the molecular subtypes of GC using archival tissues with different platforms. We filtered out genes from the epithelial-to-mesenchymal transition (EMT) and microsatellite instability-high (MSI) signatures (coefficient ≤ 0.4) followed by the ACRG molecular subtype strategy. Overall accuracy of reproduction of ACRG subtype was 66% (33/50). Given the importance of EMT subtype in future clinical trials, we further developed the minimum number of genes (10 genes) for EMT signatures correlating highly with the original EMT signatures (correlation ≥ 0.65). Using our 10-gene model, we could classify EMT subtypes with high sensitivity (0.9576) and specificity (0.811). In conclusion, we reproduced ACRG GC subtypes using different platforms and could predict EMT subtypes with 10 genes and are now planning to use them in our prospective clinical study of precision oncology in GC.