Which fingerprint in LibMCS? MinimumCS?

User 6b694845c2

20-05-2015 11:32:31

Dear ChemAxon,


I would like to ask what is the fingerprint that is being used in LibMCS for clustering?


Also on LibMCS, I would to ask if there is a MinimumCS as a preset, no matter what you can set as a MaximumCS?


Many Thanks,


Terry

ChemAxon d51151248d

21-05-2015 15:07:57

Dear Terry, 


The libmcs is a structure-based clustering tool, so it is basically does not use fingerprints for clustering. However, internally it uses the Chemical Hashed Fingerprints to speed up the structural search. 


Please see the following manual for detailed description of the libmcs:


https://docs.chemaxon.com/display/jklustor/Library+MCS+%28LibMCS%29+clustering


Best regards, 


Daniel

User 6b694845c2

01-06-2015 12:28:25

Hi Daniel,


 


Thank you very much!


I still do not understand the difference between the hashed fingerprints and fingerprints?


 


Regards

ChemAxon d51151248d

01-06-2015 12:43:10

Dear Terry, 


We have many fingerprints to use, Chemical Hashed Fingerpint is one of them. The Chemical Hashed Fingerprint is basically a linear fingerprint that uses hashing to create the binary representation of the fingerprint. You can find a very detailed explanation of how we generate it in the following manual:


https://docs.chemaxon.com/display/CD/Chemical+Hashed+Fingerprint


I hope this helps to clarify your problem. Feel free to ask if not.


Daniel 

User 6b694845c2

02-06-2015 15:51:14

Dear Daniel,


Thank you very much for the information, my understanding is better now!


I also have another question!


I did my clustering with LibMCS and set the minimum component substructure to 9, but two of my clusters have been created with a common substructure of 7 atoms



I would like to ask if there is an explanations for this as it is confusing.? Normally these three molecules should be singletons.


Regards


Terry

ChemAxon d51151248d

03-06-2015 14:43:35

Hi Terry, 


Can you send us the data necessary for us to reproduce this issue? The molecule set, and the settings of the clustering. We would need further data to investigate this issue. 


Thanks, 


Daniel