The persistent global impact of COVID-19 highlights the urgent need for integrative strategies to discover effective therapeutics. In this study, a systems biology framework was applied to elucidate the molecular landscape of COVID-19 and prioritize repositioned drug candidates. Transcriptomic data from four independent GEO datasets (GSE177477, GSE213313, GSE164805, GSE227341) were analyzed, identifying 465 differentially expressed genes (DEGs) commonly altered in patients. Functional enrichment analysis revealed key immune and inflammatory pathways dysregulations, including T cell activation and IL-17/TNF signaling. A protein–protein interaction (PPI) network constructed from these DEGs yielded 12 central hub proteins. To construct a comprehensive therapeutic target panel, these 12 network-derived hubs were integrated with three proteins well-established as indispensable for SARS-CoV-2 viral entry and processing: ACE2, TMPRSS2, and Furin, creating a 15-protein molecular signature. Principal component analysis (PCA) demonstrated the diagnostic relevance of this network signature across datasets, with TRIM25 and HIST2H2BE consistently acting as major contributors to patient stratification. Through computational drug repositioning using the L1000CDS² platform, six candidate therapeutics were identified—sirolimus, mocetinostat, simvastatin, radicicol, fostamatinib, and mitoxantrone. Molecular docking simulations further indicated that sirolimus and mocetinostat display high-affinity, multi-target binding profiles across most identified proteins, including TRIM25, HSP90AA1, APP, PRC1, OGT, HIST1H2BG, AES, YWHAH, ITGA4, CCR1, DPP4, and HIST2H2BE, outperforming reported inhibitors. To validate these static predictions, molecular dynamics (MD) simulations were performed, confirming that both sirolimus and mocetinostat maintain robust structural stability and persistent binding within the active sites of critical targets HIST2H2BE and TRIM25 under physiological-like conditions. This integrative approach provides mechanistic insights into COVID-19 pathogenesis and identifies two promising repurposed therapeutics targeting multiple disease-associated proteins reported above. The findings support using transcriptome-guided network pharmacology combined with dynamic structural validation for rational drug discovery in complex infectious diseases. Experimental validation will be essential to confirm therapeutic efficacy and clinical applicability.