Construction Of A Database Of β - Lactam Target Protein Structures For Automatic Reverse Docking

Beta-lactam antibiotic is one of the most widely used drugs in clinical treatment, its antibacterial mechanism is the inhibition of cell wall stick peptide synthetase, namely the penicillin binding protein (PBPs). Due to the overuse of antibiotics, mutants of PBPs appear ceaselessly, making the bacteria resistant to antibiotic. In the clinical treatment of bacterial infections, if the first choose of antibiotics is not sensitive enough, it may lead drug-resistant In a short time. So choosing the right first antibiotic has a great influence on the cure rate. The effect of antibiotics on bacterial strain is decided by several aspects, one of the most important is the affinity of bacteria and antibiotics. Hense, this article devote to find a way to predict the affinities of beta-lactam antibiotic and different beta lactam antibiotic target proteins quickly, accuratly, and cheaply by bioinformatics methods。In this article,56 kinds of PDB format,which are structure data of penicillin binding proteins and other related enzyme protein and some mutants of these proteins from e.coli, acinetobacter baumannii, klebsiella pneumoniae, streptococcus pneumoniae, staphylococcus aureus, MRSA, pseudomonas aeruginosa, pseudomonas aeruginosa, haemophilus influenzae and other common pathogenic bacteria were downloaded from RSCB PDB website (http://www.rcsb.org/pdb).We collect the data and use Autodock tools software to prepare these data as receptors for docking and analysis the active sites of them. Then we save these data of active sites as docking configs in TXT format, and save the prepared protein data in pdbqt format with the configs to finish the building of the beta-lactam antibiotic target protein database. At last, a script was written by MS-DOS language to make the Autodock vina software could call the database for invo-docking.After the building of database, we used 2 different types of beta-lactam antibiotics, imipenem and penicillin G, to test the accuracy of the beta-lactam antibiotic target protein database.the results showed that the prediction of the effect of imipenem was accurate, but the prediction of the effect of penicillin G was not so accurate. There were some docking results showd high affinity but the clinical effect is poorer. It was because the inv-docking could only predict the affinity of the beta-lactam antibiotic target proteins and beta-lactam antibiotics, but there are other factors affect the clinical effect of the antibiotics. So, it need users have rich clinical experience, also know a variety of bacterial resistance mechanisms.Comprehensive the above results, we think this topic has built a database of beta-lactam antibiotics target protein structure. A MS-DOS script make it could be called by Autodock vina software for auto inv-docking. This could predict the affinity of beta-lactam antibiotic target proteins and beta-lactam antibiotics at the molecular level accurately and quickly. This database has certain reference function on clinical medicine.