PHARMACEUTICAL CHEMISTRY

Author: VIDAL MONTULL DAVID.
Year: 2005.
University: BARCELONA [www.ub.es].
Place of defense: FACULTAT DE QUÍMICA.
Place of preparation: UNIVERSITAT DE BARCELONA.

Summary: The discovery and development of drugs is an interdisciplinary process and costly, both in economic terms as temporary, in which pharmaceutical companies can spend, on average, 880 million dollars and 15 years in research until a drug gets into the market. In this context, the computational methodologies can offer some advantages significant decline in 130 million dollars and 0.8 years on average, the cost of the process of discovery and drug development. Methods LINGO methodologies are highly efficient computational analysis based on the analysis of the codes SMILES, a textual representation molecular highly compressed, which based on the use of a lot of information implicitly contain the same information as any other representation based on a table of connectivity. LINGO The methods are based on a protocol of fragmentation of codes SMILES, which has allowed the development of a method for analyzing the similarity intermolecular, LINGOsim, which allows for 75,000 comparisons per second. In the same way, the methods LINGO have allowed the generation of models QSRP called LINGO-QSPR, for the prediction of molecular properties, including logP, logS or PSA, which enables the processing of more than 120,000 molecules per second. The accuracy of these models is comparable to the best methods presently available, but its speed of calculation beyond, in many cases, by more than two orders of magnitude. Moreover, it has developed an automated protocol to the practical use of the program docking Autodock 3.0. This protocol includes a large variety of programs, developed in our group as trade, which is responsible for carrying out an automated all the necessary tasks, generating structure 3D of the ligands, carrying out calculations and analysis. Thus, due to its full automation, this protocol opens the door to the analysis, through a program of docking, large quantities of molecules. The development of methods LINGO and design of a protocol automated docking, has enabled the design of an algorithm integrated virtual screening combines the use of methods of docking, Autodock 3.0, and similar molecular LINGOsim. This algorithm, which has been known as MPA (Massive Processing Algorithm) has shown its efficiency in the analysis of large databases, being able to identify the 60% of the most promising molecules of a database of more than 900,000 molecules, having done only 60,000 of docking on them, or 6.6% of the database. The integration of all of these tools has led to the development of a screening protocol virtual been validated experimentally in a draft discovery of a protein inhibitor of tyrosine fosfatas low molecular weight (ImwPTP).