Best practices for chemical data curation and QSAR model ...
Best practices for chemical data curation and QSAR model ...
Best practices for chemical data curation and QSAR model ...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Database Mining: Similarity Search vs. <strong>QSAR</strong> Search<br />
A Large Commercial Database<br />
of 515,000 Compounds<br />
Similarity Search<br />
• Similarity Metric: Tanimoto<br />
Coefficient; of every single<br />
compound in the training set<br />
• Fingerprint: MACCS Structural<br />
Keys<br />
• 425 hits obtained <strong>for</strong> TC=0.80;<br />
2 hits obtained <strong>for</strong> TC=0.90<br />
<strong>QSAR</strong> Database Search<br />
• Global search based on the<br />
whole <strong>chemical</strong> space (MZ 4.09<br />
des.) of training set<br />
• 12 hits obtained after global<br />
search (Z = 0.5) <strong>and</strong> subjected to<br />
consensus predictions<br />
• 2 selected <strong>for</strong> experimental<br />
validation based on high<br />
predicted activity, uniqueness of<br />
structure & availability<br />
There was NO overlap between the hits from two protocols; All 12<br />
<strong>QSAR</strong> hits were below TC=0.80 of training set.
Change language