PharmaBLAST: A BLAST source for drug researchers

Editing: Added "together with the 3-D map of the binding sites into the database. " at the last sentence of paragraph 3.

The good ol' NCBI (National Center for Biotechnology Information) is a quintessential element in biotechnology application and biomedical research, with its BLAST software packages comprising BLASTN, BLASTP, TBLASTX, etc, a rather ubiquitous tool.

I figure that there should be an online Bioinformatics tool that allows toxicologists and drug developers to ascertain whether the xenobiotics they have in mind will interact with important proteins. There is also a worry that drugs can interact with proteins to form hapten complexes and trigger an immune reaction, and this results in hepatitis (Anderson et al)

The process can begin when the resolved protein structures are being submitted to PDB. The proteins can be scanned for potential drug binding sites using an algorithm like PASS (Putative Active Sites with Spheres). The PASS algorithm is designed to fill putative cavities within the protein with spheres (Brady and Stouten) . An improved algorithm would be to fill the protein cavities with spheres of different properties, like the polar spheres, e.g. O, N, etc, or hydrophobic spheres. Hydrophobic spheres would more favor hydrophobic pockets within the protein, while polar spheres would favor the outer cavities of the protein with hydrophilic residues. Subsequent rounds of spheres filling will elucidate the spherical structure of the candidate xenobiotic. The last step would be to extract the spherical cavity filling structure by filtering out the rest of the protein structure and store this putative spherical drug structure, together with the 3-D map of the binding sites into the database.

When the researcher wants to test the xenobiotic in question, all he needs is to submit the structure of the xenobiotic, and the molecular spheres of the xenobiotic can be determined. The molecular sphere of the xenobiotic is then compared with the various stored spherical structures that has been earlier determined to fit into the proteins' cavities in the database and their level of similarities can be determined.

I believe having such a BLAST resource would be useful for drug developers and toxicologists, who will get to know first hand the type of proteins that their candidate xenobiotic interacts with, allowing them to elucidate the possible side effects at the cellular level. It will be good if such a database is also connected to a pathway database like BIOCARTA or STKE, and the full range of affected pathways can be determined immediately.

I would believe such a resource would be useful in Singapore with a rich biodiversity of flora and fauna that provides a wealth of potential drug candidates.

Citations
G. Patrick Brady, Jr. and Pieter F.W. Stouten. Fast Prediction and Visualization of Protein Binding Pockets with PASS. http://www.ccl.net/cca/software/UNIX/pass/pass_jcamd.html

Anderson JS, Rose NR, Martin JL, Eger EI, Njoku DB. Desflurane hepatitis associated with hapten and autoantigen-specific IgG4 antibodies. Anesth Analg. 2007 Jun;104(6):1452-3.