Objective: To take technology of computer virtual screening for fast searching small molecule inhibitors for H7N9 subtype of influenza virus neuraminidase(NA) from traditional Chinese medicine database@Taiwan(TCM database@Taiwan). Method: Based on optimized complex structure of NA bound with specific inhibitor of Zanamivir

computer-aided structure-based virtual screening against TCMD database@taiwan was conducted to determine occurrence of herb-based NA inbitors in AutoDock Vina software package of Scripps institution.Virtual screening results were further filtered by predictive ADME simulation by traditional Chinese medicine systems pharmacology database and analysis platform(TCMSP) and simultaneously filtered by predictive toxic simulation using TOPKAT module from Discovery Studio 2.5 software package. Result: Free binding energy of original ligand(Zanamivir) could be used as a threshold

then not only obtained three compounds having a higher binding affinity than Zanamivir

but also they had good drug-likeness.Their sources of traditonal Chinese medicine had been determined. Conclusion: This study provides an important reference and a theoretical basis for extraction of antiviral compounds from Chinese herbal medicine and design of anti-influenza drugs.