Web Search of New Linearized Medical Drug Leads
This addresses the challenge for researchers in drug discovery by enabling more effective search of multi-dimensional drug information, though it is incremental as it adapts existing search methods.
The authors tackled the problem of searching for medical drug leads on the Web by using linearized molecular structure fragments as inputs to overcome the limitation of keyword-only search engines, resulting in manageable and rich result sets with relatively small ambiguity.
The Web is a potentially huge source of medical drug leads. But despite the significant amount of multi- dimensional information about drugs, currently commercial search engines accept only linear keyword strings as inputs. This work uses linearized fragments of molecular structures as knowledge representation units to serve as inputs to search engines. It is shown that quite arbitrary fragments are surprisingly free of ambiguity, obtaining relatively small result sets, which are both manageable and rich in novel potential drug leads.