Technology Q&A
- What is enabled by our technology?
The representation of chemical compounds as graphs captures the essential aspects of chemical structures in a natural and flexible way. The application of recent techniques for graph querying and mining has demonstrated scalability as well as an improved quality of results over traditional representation techniques such as fingerprints. Our proprietary methods include novel ways of graph comparison, the organization of graphs in a hierarchical index structure, and the mining of a set of graphs to find statistically over-represented motifs.
- How is SimFinder different from fingerprint-based similarity search?
The connectivity of the chemical functional groups, lost in fingerprints, is maintained. Additionally, because the overlap between topology-based and fingerprint-based similarity search results is small, topology-based queries provide new insight into previously mined chemical libraries.
- How have we validated your technology?
We have performed retrospective studies to enrich compound libraries of known biological activity. For example, similarity search queries with known HIV-1 Protease inhibitors returns HIV-1 PR active compounds. Additionally, we have conducted reviews of our search results with Medicinal Chemists. We have applied GraphSig to the classification of chemical compounds and the results were better than finger prints. Additionally, GraphSig was able to extract meaningful substructures from collections of chemical compounds.