From: Prateek G. <pra...@gm...> - 2013-04-24 17:09:51
|
hi, sorry for the inconvinience for my last post not on the mailing list. By my last post I was trying to convey that which I have understood and to seek your opinion. I used the version 1.7 because it was pointed out in a previous mail.I will install the 2.x version soon. I have read the CT-index research paper and what i have learnt uptil now. By the subset reason I was trying to convey about the degree of fingerprint matching of the features of the query pattern with the features of the database graph pattern.Those that were positive in the test were used for the isomorphism procedure and discarding the rest. This is the string matching algorithm that is being used in the formal definition of certificate matching? So I aim to know what exactly does the fragment z value that is currently limited to 0-6 accounts for in the filter rules .Is it the number of features to be taken into account while filtering or the number of atoms that must be possessed in the drug compound , of the query atom that filter mechanism is considering or is it related to the fingerprinting bit array? Also does this project idea aims to integrate the JChemPaint as the new editor over the current structure editor being used. Regards, Prateek |