BDL-SP: A Bio-inspired DL model for the identification of altered Signaling Pathways in Multiple Myeloma using WES data

Identification of the genomic features responsible for the progression of Multiple Myeloma (MM) cancer from its precancerous stage MGUS can improve the understanding of the disease pathogenesis and, in devising suitable preventive and treatment measures. We have designed an innovative AI-based model, namely, the Bioinspired Deep Learning architecture for the identification of altered Signaling Pathways (BDL-SP) to discover pivotal genomic biomarkers that can potentially distinguish MM from MGUS. The proposed BDL-SP model comprehends gene-gene interactions using the PPI network and analyzes genomic features using a deep learning (DL) architecture to identify significantly altered genes and signaling pathways in MM and MGUS. For this, whole exome sequencing data of 1174 MM and 61 MGUS patients were analyzed. In the quantitative benchmarking with the other popular machine learning models, BDL-SP performed almost similar to the two other best performing predictive ML models of Random Forest and CatBoost. However, an extensive post-hoc explainability analysis, capturing the application specific nuances, clearly established the significance of the BDL-SP model. This analysis revealed that BDL-SP identified a maximum number of previously reported oncogenes, tumor-suppressor genes, and actionable genes of high relevance in MM as the top significantly altered genes. Further, the post-hoc analysis revealed a significant contribution of the total number of single nucleotide variants (SNVs) and genomic features associated with synonymous SNVs in disease stage classification. Finally, the pathway enrichment analysis of the top significantly altered genes showed that many cancer pathways are selectively and significantly dysregulated in MM compared to its precursor stage of MGUS, while a few that lost their significance with disease progression from MGUS to MM were actually related to the other disease types. These observations may pave the way for appropriate therapeutic interventions to halt the progression to overt MM in the future.

Automated summary

Identifying genomic features responsible for the progression of Multiple Myeloma (MM) from precancerous stage MGUS can enhance our understanding of the disease and aid in developing appropriate preventive and treatment measures. This study introduces an innovative AI-based model, called Bioinspired Deep Learning architecture for the identification of altered Signaling Pathways (BDL-SP), which effectively identifies crucial genomic biomarkers that can distinguish MM from MGUS. The evaluation demonstrates the significance of the BDL-SP model through extensive post-hoc explainability analysis and benchmarking against other machine learning models. The findings provide insights into therapeutic interventions to halt the progression to overt MM. Rating: 9/10.