Date of Final Oral Examination (Defense)
Type of Culminating Activity
Masters of Science in Computer Science
Timothy Andersen, Ph.D.
Owen McDougal, Ph.D.
Amit Jain, Ph.D.
Since the discovery of the molecular basis of disease, numerous studies have reported a correlation between the activity of specific protein receptors the progression of disease. As a result, drug development has become dependent on the study of protein receptor activities. The relative inexpense of computing hardware has made computational methods an important supplementary tool for receptor modeling. This work details an open source software tool that is capable of both efficiently screening large peptide mutant libraries and enabling 3D conformer-based searches over local molecular databases.
A Computational Approach to Efficient Peptide Influenced Drug Repurposing (CAEPIDR) has been developed to explore the conformational ligand binding space of the α3β2 nicotinic acetylcholine receptor (nAChR) isoform using a library of α- conotoxin (α-CTx) MII peptide mutants. The screen’s top hits were used to identify small molecule drugs that might also bind to the receptor. The conformational ligand binding space of the nAChR was heuristically explored using a genetic algorithm, which managed a structure-based virtual screen of a 640,000 α-CTx MII peptide mutant library. A utility was developed to search the PubChem Compound database for small molecule drugs with a 3-D shape similar to the highest affinity peptides from the virtual screen.
CAEPIDR’s genetic algorithm based procedure was able to find 10 peptides with estimated free energies of binding (with the α3β2-nAChR) below -20 kcal/mol, which can be compared to α-CTx MII’s -12.38 kcal/mol. These peptide’s were identified in spite of the genetic algorithm performing docking calculations for only 9344 of the 640,000 α-CTx MII mutants. The PubChem Compound search yielded 2 small molecule drugs with estimated binding energies below -20 kcal/mol.
CAEPIDR has been integrated with DockoMatic to create DockoMatic 2.1, which can be used to create virtual peptide mutant libraries, virtually dock ligands to macromolecular receptors, and identify small molecule drugs for disease treatment.
Long, Thomas Francis, "CAEPIDR: A Computational Approach to Efficient Peptide Influenced Drug Repurposing" (2014). Boise State University Theses and Dissertations. 877.