Integrative molecular network analysis of genetic risk factors to infer biomarkers and therapeutic targets for rheumatoid arthritis

Abstract

Background Understanding the interplay between genetic risk factors and molecular pathways in rheumatoid arthritis (RA) is essential for developing effective treatments. This study aims to utilize genetic risk factors of RA and identify their key pathways and potential therapeutic targets through an integrated multi-omics approach. Methods We developed a computational pipeline to construct a knowledge graph that combines genetic risk factor molecular networks with multi-omics enrichment analysis to estimate potential therapeutic target for RA. Genetic risk factors were extracted from the literature, curated, and annotated. Molecular interaction networks were constructed based on these genetic risk factors and their neighboring proteins. Enrichment analysis was performed to identify significantly impacted biological processes and pathways. Multi-omics knowledge graph was used to prioritize candidates potential therapeutic target for RA. Results Our analysis identified 35 significant genes associated with RA as potential therapeutic targets and biomarkers, categorized into three pathways: Cytokine Regulation and Production, Hematopoietic or Lymphoid Organ Development, and Myeloid Cell Differentiation. Among these, 25 genes were classified as risk genes, while 10 were neighboring genes. We identified nine novel risk proteins (RELA, ETS1, NFATC1, BATF, LCK, PIK3R1, PRKCB, RASGRP1,and FYN) as potential therapeutic targets for RA and they significantly contribute in the disease pathogenesis. Conclusion This study provides a comprehensive integrative molecular network and knowledge graph analysis of genetic risk factors in RA, offering a solid framework for integrating multi-omics data in RA research. These findings may contribute to more accurate clinical decision-making and the development of targeted treatment regimens. Additionally, this study highlights the importance of inferring hidden relationships across networks based on disease associations and functional similarities, further enhancing our understanding of RA pathogenesis.