Human respiratory syncytial virus (hRSV) affects over 33 million people each year, yet currently there are no approved effective drugs or vaccines. In this study, we first constructed a candidate host-pathogen species-wide genetic and epigenetic network (HPI-GWGEN) through big data mining. Subsequently, we trimmed false positives in the candidate HPI-GWGEN to obtain the true HPI-GWGEN by analyzing time-resolved dual host-pathogen RNA-seq data and applying systems biology approaches. Utilizing main network projection and KEGG pathway annotation, we were able to extract core signaling pathways during hRSV infection and investigate the pathogenic mechanisms of hRSV infection, selecting important biomarkers as drug targets, namely TRAF6, STAT3, IRF3, TYK2, and MAVS. Finally, to discover potential molecular drugs, we trained a deep neural network-based DTI prediction model using drug-target interaction (DTI) databases to predict candidate molecular drugs for these drug targets. Simultaneously, we screened these candidate molecular drugs based on three drug design criteria, namely control ability, sensitivity, and toxicity. Ultimately, we selected Arbidol, RS-67333, and Phenformin as potential multi-molecular drugs for treating hRSV infections.