Identifying Drug Targets in Pancreatic Ductal Adenocarcinoma Through Machine Learning, Analyzing Biomolecular Networks, and Structural Modeling

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is one of the leading causes of cancer-related death and has an extremely poor prognosis. Thus, identifying new disease-associated genes and targets for PDAC diagnosis and therapy is urgently needed. This requires investigations into the underlying molecular mechanisms of PDAC at both the systems and molecular levels. Herein, we developed a computational method of predicting cancer genes and anticancer drug targets that combined three independent expression microarray datasets of PDAC patients and protein-protein interaction data. First, Support Vector Machine-Recursive Feature Elimination was applied to the gene expression data to rank the differentially expressed genes (DEGs) between PDAC patients and controls. Then, protein-protein interaction networks were constructed based on the DEGs, and a new score comprising gene expression and network topological information was proposed to identify cancer genes. Finally, these genes were validated by "druggability" prediction, survival and common network analysis, and functional enrichment analysis. Furthermore, two integrins were screened to investigate their structures and dynamics as potential drug targets for PDAC. Collectively, 17 disease genes and some stroma-related pathways including extracellular matrix-receptor interactions were predicted to be potential drug targets and important pathways for treating PDAC. The protein-drug interactions and hinge sites predication of ITGAV and ITGA2 suggest potential drug binding residues in the Thigh domain. These findings provide new possibilities for targeted therapeutic interventions in PDAC, which may have further applications in other cancer types.

Comprehensive analysis of abnormal expression, prognostic value and oncogenic role of the hub gene FN1 in pancreatic ductal adenocarcinoma via bioinformatic analysis and in vitro experiments

“…ITGA2 has been reported to play a critical role in modulating the pancreatic cancer immune response by transcriptionally increasing the expression of PD-L1 in cancer cells (Ren et al, 2019). At the same time, it has been reported that ITGA2 is associated with unfavorable survival of PDAC patients by affecting the malignant biological behavior of pancreatic cancer cells (Rozengurt, Sinnett-Smith & Eibl, 2018;Shimomura et al, 2020;Yan et al, 2020). Considering that the mechanism of ITGA2 in PDAC has been studied partly, we focused on FN1 which has no mechanism study in PDAC for further study.…”

Identification of the Neurokinin-1 Receptor as Targetable Stratification Factor for Drug Repurposing in Pancreatic Cancer

“…So far, only a very small number of studies has investigated the effects of AP in cancer patients on a molecular level regarding its role as a potential cancer drug. Emphasizing on PDAC being a very heterogenic tumor with no dominant druggable mutation [ 46 , 47 ], investigating the mechanisms on a transcriptional level has high potential in highlighting crucial differences between tumor subtypes and their underlying mechanisms in tumor progression. With no substantial improvements in PDAC treatments over the past 30 years [ 48 , 49 ], such information has the power to lead to new therapeutic developments.…”

Centrality of drug targets in protein networks

“…Recent analyses evaluated network characteristics of drug targets combined with annotated functional data in disease specific contexts [50,51]. With the goal of identifying suitable drug targets in Pancreatic Ductal Adenocarcinoma, Yan et al devised a 'hybrid' RNs score ranking that combined information from gene expression datasets with node centrality metrics (average shortest path length, degree) within a sub-network of the String database relevant to this disease [50]. Kim et al identified a disease-relevant protein network (module) for Systemic Sclerosis and evaluated it with enrichment analysis of Gene Ontology (GO) biological processes descriptors [51].…”

“…S16c, d and Table S7), possibly convoluting the differences in disease associations between Phase4 and all targets with a consequential rather than causal relationship to the fact that drug targets are extensively studied proteins, as discussed at length previously. We combined the percentile ranking of disease associations with graph centrality metrics, similarly to the cited RNs score [50] (with the difference that the original RNs score was derived from gene expression data rather than annotated disease associations) and with centrality metrics identified in Clustering, present in the Watts Strogatz (WS) random network (e), disrupts the correlation between the two parameters with a random noise effect. For the target class of enzymes (f), Phase4 targets exhibit significantly lower topological coefficient than all enzymes in String0.7 even if the average degree of Phase4 enzymes is lower.…”