I think this MCS work I did last night is pretty neat. I got the list of structures in every ChEBI structure ontology node, found the MCS for the set, and generated a (very large) visualization of the entire results set.
It's at http://dalkescientific.com/fmcs_chebi.html.bz2 . The 7.5 MB file uncompresses to 166 MB, in part because I wasn't concerned about space when I wrote the code to integrate with the SMARTSViewer and Daylight image services. Some background. Last evening I downloaded the most recent ChEBI chemical ontology. This organizes structures into one or more families, and organizes the families in turn into higher level families. The result is a hierarchy. For example, CHEBI:33567 is catecholamine, which includes as children: CHEBI:37950 hexoprenaline CHEBI:50580 arbutamine CHEBI:6257 L-isoprenaline CHEBI:33568 adrenaline CHEBI:33569 noradrenaline CHEBI:4670 dobutamine CHEBI:18243 dopamine It's in turn a child of the following set of relations: CHEBI:33822 organic hydroxy compound CHEBI:23824 diol CHEBI:22625 aromatic diol CHEBI:33570 benzenediols CHEBI:33566 catechols You can see its full hierarchy in http://www.ebi.ac.uk/chebi/chebiOntology.do?treeView=true&chebiId=CHEBI:33567#graphView My question is, how well does my MCS code work when given one of the intermediate nodes and the list of structures which make up the node? Note: not all of these have a simple substructure scaffold which my MCS code can detect. Their paper at http://www.biomedcentral.com/1471-2105/13/3 points out that certain patterns, like ester, require a SMARTS to match C(=[O,S])OC and that some terms, like bicyclic molecules, cannot be captured via a simple pattern. It took about 50 minutes to process the MCSes, once I fixed a bug in my code, and several hours to understand the EBI data format and to make a relatively interesting visualization. I hope you enjoy exploring the result - let me know if there's anything interesting you find! Andrew [email protected] ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Rdkit-discuss mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
