Identifying multi-omics interactions for lung cancer drug targets discovery using Kernel Machine Regression

TL;DR

Kernel Machine Regression (KMR) is applied to integrate three lung cancer omics layers—gene expression, DNA methylation, and miRNA—to uncover higher-order interactions and identify potential drug targets. The approach uses an RKHS-based multi-view model with variance-component testing for marginal, interaction, and composite effects, complemented by LIMMA/T-test/Wilcoxon/CCA-based preselection and molecular docking for drug repurposing. The analysis yields 9 significant triplets across 5 genes, 7 transcripts, and 3 epigenomes, plus 38 associated genes and eight hub genes (PDGFRA, ITGB1, SNAI1, FGF11, PDGFRB, ID1, TNXB, ZIC1), with Selinexor, Orapred, and Capmatinib emerging as top candidates. This work demonstrates a robust framework for non-linear, multi-omics integration to inform targeted therapy in lung cancer, while emphasizing the need for high-quality input data for reliable results.

Abstract

Cancer exhibits diverse and complex phenotypes driven by multifaceted molecular interactions. Recent biomedical research has emphasized the comprehensive study of such diseases by integrating multi-omics datasets (genome, proteome, transcriptome, epigenome). This approach provides an efficient method for identifying genetic variants associated with cancer and offers a deeper understanding of how the disease develops and spreads. However, it is challenging to comprehend complex interactions among the features of multi-omics datasets compared to single omics. In this paper, we analyze lung cancer multi-omics datasets from The Cancer Genome Atlas (TCGA). Using four statistical methods, LIMMA, the T test, Canonical Correlation Analysis (CCA), and the Wilcoxon test, we identified differentially expressed genes across gene expression, DNA methylation, and miRNA expression data. We then integrated these multi-omics data using the Kernel Machine Regression (KMR) approach. Our findings reveal significant interactions among the three omics: gene expression, miRNA expression, and DNA methylation in lung cancer. From our data analysis, we identified 38 genes significantly associated with lung cancer. From our data analysis, we identified 38 genes significantly associated with lung cancer. Among these, eight genes of highest ranking (PDGFRB, PDGFRA, SNAI1, ID1, FGF11, TNXB, ITGB1, ZIC1) were highlighted by rigorous statistical analysis. Furthermore, in silico studies identified three top-ranked potential candidate drugs (Selinexor, Orapred, and Capmatinib) that could play a crucial role in the treatment of lung cancer. These proposed drugs are also supported by the findings of other independent studies, which underscore their potential efficacy in the fight against lung cancer.

Figures (6)

• Figure 1: The impact of multi-omics data interaction on human lung tissue. The figure displays the gene expression (GE), micro ribonucleic acid (miRNA), and deoxyribonucleic acid methylations (DNA) data for human lung tissue, and highlights the potential interactions between these omics layers.
• Figure 2: Differential gene expression analysis using (a) T-test, (b) LIMMA, (c) Wilcoxon, (d) CCA. See text for thresholds.
• Figure 3: Venn diagrams for gene selection overlap across T-test, LIMMA, Wilcoxon, and CCA for (a) gene expression, (b) DNA methylation, (c) miRNA.
• Figure 4: Manhattan plots of $-\log_{10}(p)$ versus triplets based on KMR overall tests.
• Figure 5: Venn diagram of hub genes across topological algorithms (MCC, MNC, Radiality, EPC, Degree).