Identifying potential drug targets in hepatocellular carcinoma based on network analysis and one-class support vector...

Area of Science:

• Oncology
• Bioinformatics
• Computational Biology

Background:

• Hepatocellular carcinoma (HCC) is a significant global cause of cancer mortality.
• Current systemic therapies for advanced HCC are limited, necessitating novel therapeutic strategies.
• Identifying effective drug targets is crucial for advancing HCC treatment.

Purpose of the Study:

• To develop an in-silico drug target predictor for hepatocellular carcinoma.
• To leverage clinical data, gene expression, and protein-protein interaction networks for target identification.

Main Methods:

• Integrated clinical association data, gene expression profiles, and known drug target genes with a human protein-protein interaction network.
• Analyzed network properties (degree, centrality) and genetic dependency scores of various gene sets (DTGs, DAGs, PUGs, URGs, PFGs, DRGs).
• Constructed a one-class support vector machine (one-class SVM) model based on identified features of drug target genes.

Main Results:

• Drug target genes (DTGs) and other disease-related genes exhibited distinct network properties compared to background genes.
• DTGs showed closer proximity to disease-associated and unfavorable prognostic genes within the network.
• The developed one-class SVM predictor demonstrated effectiveness in identifying potential HCC drug targets.

Conclusions:

• The study successfully established an in-silico predictor for identifying novel drug targets in HCC.
• The findings provide a valuable tool for accelerating the discovery of targeted therapies for HCC.
• This computational approach can guide future research in developing more effective treatments for advanced liver cancer.