I am conducting a master's research project where I have incorporated a terpene into a polysaccharide-based hydrogel and aim to evaluate its osteoinductive activity in mesenchymal stem cells using molecular biology techniques. To complement this, I am performing a network pharmacology analysis to explore potential targets of the terpene that may be involved in osteogenesis.
Here is the workflow I have followed so far:
Identified terpene targets using SwissTargetPrediction and osteogenesis-related genes via GeneCards.Filtered and intersected the results through a Venn diagram to identify common targets.Imported the common targets into STRING and exported the TSV file for PPI network analysis in Cytoscape.Now, I am exploring the most robust approach for further analysis and would appreciate insights on the following options:
Perform GO and KEGG enrichment analysis on all common targets identified.Analyze the PPI network in Cytoscape, calculate metrics such as degree and closeness, and select hub genes based on values above the median or a fixed threshold (e.g., top 10, 20, or 30 genes). Then, conduct GO and KEGG analysis on these hub genes.Use CytoHubba (e.g., MCC criterion) to select hub genes and perform enrichment analysis.Cluster the network (using methods such as MCODE or MCL) and evaluate GO and KEGG enrichment for each cluster. Which clustering method would be most appropriate in this context?If both hub gene analysis and cluster analysis are performed, how should I integrate these results? Should clusters be selected based on size or another criterion?I would greatly value your suggestions on which approach would yield the most meaningful biological insights or if there are alternative methods I should consider. Thank you in advance for your guidance!