SPIDR: small-molecule peptide-influenced drug repurposing

• Published in: BMC bioinformatics 2018-04, Vol.19 (1), p.138-138, Article 138
• Main Authors: King, Matthew D, Long, Thomas, Pfalmer, Daniel L, Andersen, Timothy L, McDougal, Owen M
• Format: Article
• Language: English
• Subjects: 60 APPLIED LIFE SCIENCES; Biochemistry & Molecular Biology; Biotechnology & Applied Microbiology; DockoMatic; Drug discovery; Drug repurposing; GAMPMS; Innovations; Mathematical & Computational Biology; Methods; Repositioning; SimSearcher; Therapeutics research
• ISSN: 1471-2105; 1471-2105
• DOI: 10.1186/s12859-018-2153-y

Conventional de novo drug design is costly and time consuming, making it accessible to only the best resourced research organizations. An emergent approach to new drug development is drug repurposing, in which compounds that have already gone through some level of clinical testing are examined for efficacy against diseases divergent than their original application. Repurposing of existing drugs circumvents the time and considerable cost of early stages of drug development, and can be accelerated by using software to screen existing chemical databases to identify suitable drug candidates. Small-molecule Peptide-Influenced Drug Repurposing (SPIDR) was developed to identify small molecule drugs that target a specific receptor by exploring the conformational binding space of peptide ligands. SPIDR was tested using the potent and selective 16-amino acid peptide α-conotoxin MII ligand and the α β -nicotinic acetylcholine receptor (nAChR) isoform. SPIDR incorporates a genetic algorithm-based, heuristic search procedure, which was used to explore the ligand binding domain of the α β -nAChR isoform using a library consisting of 640,000 α-conotoxin MII peptide analogs. The peptides that exhibited the highest affinity for α β -nAChR were used as models for a small-molecule structure similarity search of the PubChem Compound database. SPIDR incorporates the SimSearcher utility, which generates shape distribution signatures of molecules and employs multi-level K-means clustering to insure fast database queries. SPIDR identified non-peptide drugs with estimated binding affinities nearly double that of the native α-conotoxin MII peptide. SPIDR has been generalized and integrated into DockoMatic v 2.1. This software contains an intuitive graphical interface for peptide mutant screening workflow and facilitates mapping, clustering, and searching of local molecular databases, making DockoMatic a valuable tool for researchers in drug design and repurposing.