Poster #RP212
Towards a 3D Structure Database of Therapeutic Targets for Inverse Docking
Julio Lopes*, Andrelly Jose*, Sergio Alencar*
*Chemoinformatics and Medicinal Chemistry Group, Departamento de Quimica - UFMG, Belo Horizonte, Brazil
The great demand on improving the efficiency of drug discovery has created a need for a new paradigm that enables more scientists to use structural information in the combinatorial chemistry and medicinal chemistry process. Recently, an innovative approach called inverse docking had been introduced as a chemoinformatics tool to predict several characteristics of lead candidates. From a database of biological and therapeutic targets, one could (i) identify unknown and secondary therapeutic targets for a drug, (ii) predict the potential toxicity and side effects of an investigating drug, and (iii) probe the molecular mechanism of action of bioactive compounds.
We used the drug targets sequences from Drugbank (http://redpoll.pharmacy.ualberta.ca/drugbank/) to search for PDB sequences which they are related. The NCBI Blast was used to compare the sequences and to cluster PDB sequences (12205 clusters) and targets sequences (4042 clusters) with a 90% identity threshold. We selected 1977 PDB structures with high resolution (less than 2.5 A) and 100% identity with targets sequences to populate our therapeutic target database. We also included the Interpro and Swiss-Prot entries related to PDB chains to obtain the associated biological activities. Now, we are working to develop a website with the 3D structures of therapeutic targets (http://www.nequim.qui.ufmg.br).