In recent years, outbreaks of avian influenza rapidly spread around the world. Avian influenza (AI) was caused by Type-A influenza virus. According to viral sur-face glycoprotein, there have been identified sixteen HA subtypes (H1-H16) and nine NA subtypes (N1-N9). Reviewing the revolution history of avian influenza, most of so-called highly pathogenic avian influenza (HPAI) outbreaks were originated from low virulent H5 or H7 viruses, and then mutated into high pathogenic ones. The muta-tion of influenza virus occurred such that many new virus strains were discovered each year. The recombination of different types of avian influenza virus occurs so fre-quently that human might be infected someday soon. The possible pandemic of novel influenza could cause millions of death finally. It is the first step of viral infection the binding event between surface glycopro-tein Hemagglutinin (HA) and Sialic acid on host cell surface receptors. In the litera-ture, it was shown that the binding specificity of hemagglutinin (HA) receptor of the influenza virus could be changed only by a single amino acid substitution. As a result, the change of binding specificity between hemagglutinin and its receptor may raise its pathogenicity, and then it might trigger a possible influenza pandemic. In this paper, we investigated 82 hemagglutinin sequences of H5N1 influenza vi-rus from Swiss-Prot protein database. We analyzed the evolutionary rates among the sequences for the investigation of possible locations where may cause the change of binding specificity. The specificity-determining positions (SDPs) in families of he-magglutinin proteins were studied. We identified two SDPs from temporal analysis within the years of 1996-2005, and three SDPs from spatial analysis between different regions. Moreover, we endeavored to investigate potential drugs for avian influenza. We targeted the specific sites from previous sequence analysis as possible pharmacophore. Then the sites were used to find lead compounds for pharmaceutical purpose with re-ceptor-based structural analysis. We used the docking tool, GOLD, to select potential lead compounds from the Maybridge database. After screening 59284 objects in the database, two potential molecules were selected as lead compounds. The compound, (diaminoBLAHyl)-naphthalen-4-yl-methanone, is the top molecule according to the affinity with the HA protein.
