Dear Stéphane,
I'll look into this tonight and let you know.
Best,
p.
On 11/17/16 19:09, Téletchéa Stéphane wrote:
Dear all,
I am willing to compute the rmsd of a docking pose against a
reference. My code works for some results, but not for all...
The pseudo-algorithm is:
a) read and clean the reference for the pdb file
b) read each pose (in SD format, Accelrys)
c) perform the substructure match and get the corresponding atoms
d) compute the rmsd
e) output the result.
The code and input file are attached to the message (adjust line 10
for rdkit path if needed)
The output is:
./calc_rmsd_simple.py Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamine 0.601627
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamine 0.269822
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamine 0.434981
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamine 0.454281
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamine 0.450157
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamine 0.444344
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamine 0.561584
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamine 0.526337
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamine 0.566695
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamine 0.397786
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamineTypeI 0.000000
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamineTypeI 0.000000
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamineTypeI 0.000000
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamineTypeI 0.000000
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamineTypeI 0.000000
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamineTypeI 0.000000
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamineTypeI 0.000000
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamineTypeI 0.000000
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamineTypeI 0.401164
Beta-D-Galactose_3ZSJA_H.pdb jeu_test.sd carbLactosamineTypeI 0.391936
For the first molecule the rmsd is correct, while for the second (same
input file, 10 different output poses) nearly no values can be computed.
Visually, the rmsd with 0.0 values are oriented in another in
opposition to the others.
I assumed the GetSubstructMatch(es) would be independent of the atoms
order, but it seems not. Am I using the wrong logic (or another
routine should be used, or I did not understood properly)?
I have tried to sanitize the molecules, to add / remove hydrogens, to
use various input formats (mol, mol2, sd, pdb), to reimport the
molecules into rdkit to clean them (thus the resulting test cases
above), without luck.
Thanks a lot in advance for comments and remarks, this is driving me
crazy.
Stéphane
------------------------------------------------------------------------------
_______________________________________________
Rdkit-discuss mailing list
Rdkit-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
------------------------------------------------------------------------------
_______________________________________________
Rdkit-discuss mailing list
Rdkit-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
