-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Jul 31, 2008, at 3:38 AM, Egon Willighagen wrote: > Hi all, > > I think we have been challenged by the Unilever Cambridge Center for > Molecular Informatics, Dr Jonathan Goodman, Prof Robert Glen and Dr > Antonio Llinàs [0]. > > The challenge not just us, but any data modeler, by asking us to > predict the solubility of 32 compounds, based on a computational model > we can train against 100 knowns. Now, in my experience 100 is a bit on > the low side, but at least higher than the bare limit to get low > enough statistics to distinguish prediction models. Or, in other > words, we barely can make a choice of a best model. But just enough to > pick up the challenge I assume. > > Now, any one can make a reasonable model... the real challenge to me > is to make a reproducable model, a model that anyone in the world can > build themselves. So, with Blue Obelisk components (or other > opensource tools). Reproducible is definitely a worthy model. But equally important is quality. You're right that given the small training set, it might a bit tricky to get a good quality model. But are you thinking of straight up QSAR? Recent papers [1,2] suggests that crystal lattice energies should be taken into account - but for that they used structures from the CCDC. At the very least, we'd need some sort of solvent enabled conformer generator. I don't know whether taking into account multiple confs (say via Boltzmann weighting) would be useful Also good partial charges would be useful (MOPAC would be a good option for that). It might also be useful to derive some sort of interaction descriptors from MD runs (described in [2]) [1] http://dx.doi.org/10.1021/mp7000878 [2] http://dx.doi.org/10.1021/mp070030+ - ------------------------------------------------------------------- Rajarshi Guha <[EMAIL PROTECTED]> GPG Fingerprint: D070 5427 CC5B 7938 929C DD13 66A1 922C 51E7 9E84 - ------------------------------------------------------------------- A beer delayed is a beer denied. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.8 (Darwin) iEYEARECAAYFAkiRwXAACgkQZqGSLFHnnoT3rACglLcJE8KFhSMZ3rFg3RdeEMBP /eMAn3nKpzjcbpQRC+aF3sCJu3Rz5jAo =8j9j -----END PGP SIGNATURE----- ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ Blueobelisk-discuss mailing list Blueobelisk-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/blueobelisk-discuss
