A working prototype ‘Mira’ has been designed that can query the biodegradation prediction systems MEPPS and CATABOL, and the CATABOL pathway database about a compound of interest and reason about the predictions returned. The program can add confidence where there is concordance between inputs from difference sources, and also deals with situations where there is discordance between the inputs. The results are output as a number of files which report a qualitative likelihood of whether the compound will biodegrade, how readily it will do so and what the likely degradants will be. The predictions are assigned a confidence which is based on user experience with the systems being used. The prototype demonstrates how information from qualitative and quantitative in silico methods can be used to give advice in such scenarios.
The prototype was presented at the NoMiracle meeting in Leipzig in April 2008.