I don't know if any other of the comunity especially for cemometricians knows that there is the jcompoundmapper node realized for knime . This node add lots of popular fingerprints as CATS2D, molprint2D and other.
It is possible to find at sourceforge , at this link
http://sourceforge.net/projects/jcompoundmapper/files/
Wonderful! i have been using the command line version. Let me test!
The node should still be considered as a beta version! When we consider it ready for public use we will provide an knime update side - and maybe offer/integrate it in one of the community projects. I already spoke with some of the CDK node developers at the Knime OS meeting last week, because due to its dependence on the CDK the node would probably be placed best in their Knime extension.
I'll keep you posted!
In the meantime, feel free to test and please report found bugs on the sourceforge project side.
Nikolas
Hi Nikolas
thanks for your precious work. Really nice!
Andrea
Any updates..release date for the community update?
Hi guys,
I'm a CDK lover and I'm trying to use JCompoundMapper in order to measure molecular similarity.
I have a very odd behavior here. If I calculate the fingerprint vector for a certain molecule with 2 different nodes (but same setting) I obtain 2 different fingerprint vectors. In this manner it's impossible to obtain good molecular similarity estimation. I'm using Daylight-like atom type and extended connectivity fingerprints type.
I suppose there is something wrong with the hash function. Maybe is there a random seed included there? Can you please help me with that?
Thanks,
Gio
Hi guys,
I just realized that the problem is not generalized to all the JCompoundMapper fingerprint types. For example using 2D-molprint like, this problem does not appear.
Gio