In Listing 5, we calculate the mean error for 1000 such simulations. ... This relationship between chemical similarity and model performance is detailed in a 2004 paper by Sheridan and coworkers . ... Once we havecalculated the maximum similarityofeach trainingsetstructure to eachtest setstructure, we can use aanbsp;...
|Title||:||Chemoinformatics for Drug Discovery|
|Publisher||:||John Wiley & Sons - 2013-09-25|