User 677b9c22ff
01-09-2008 07:20:59
Hi,
calculating the 3D surface area gives different values coming from either the 2D or 3D calculated structure.
I agree that this probably depends on the selected accuracy,
but the annoying or bad problem is that depending on the molecule the time or timeout values are extremely high.
On a structure around 400 Da the 3D calculation should not last longer than 10 seconds. However in certain cases (probably due to the absence of correct 3D templates) the calculation time is in the minutes range (Core2Duo 2 GHz).
My problem is not the difference in accuracy, but rather
the calculation time. Generating the 2D to 3D structure with geometry "very strict" takes around 8 seconds
(which is pretty slow), and the solvent accessible area
takes around 0.5 seconds.
Coming from 2D and calculating directly the solvent accessible area in Mview or in cxcalc the whole process takes minutes and in cxcalc no value is generated due
to a time-out error. The command would be like:
Solvent accessible area (3D) of
"COC(=O)[C@H]1C2[C@]3(C)CC[C@@H](O[Si](C)(C)C)[C@@]2(OC3=O)C4CC[C@@H]5C[C@]14CC5=C"
Both started in two independent windows with mview 5.0.1
One problem(?) is that the configuration switched, so in the two cases there are two different molecule values exported,
which is somehow weird.
methyl (1R,5R,8R,9S,11S,14R)-11-methyl-6- methylidene-16-oxo-14-[(trimethylsilyl)oxy]-15- oxapentacyclo[9.3.2.1^{5,8}.0^{1,10}.0^{2,8}]heptadecane-9- carboxylate
methyl (1S,2S,5R,8R,9S,10R,11S,14R)-11-methyl-6- methylidene-16-oxo-14-[(trimethylsilyl)oxy]-15- oxapentacyclo[9.3.2.1^{5,8}.0^{1,10}.0^{2,8}]heptadecane-9- carboxylate
Cheers
Tobias
calculating the 3D surface area gives different values coming from either the 2D or 3D calculated structure.
I agree that this probably depends on the selected accuracy,
but the annoying or bad problem is that depending on the molecule the time or timeout values are extremely high.
On a structure around 400 Da the 3D calculation should not last longer than 10 seconds. However in certain cases (probably due to the absence of correct 3D templates) the calculation time is in the minutes range (Core2Duo 2 GHz).
My problem is not the difference in accuracy, but rather
the calculation time. Generating the 2D to 3D structure with geometry "very strict" takes around 8 seconds
(which is pretty slow), and the solvent accessible area
takes around 0.5 seconds.
Coming from 2D and calculating directly the solvent accessible area in Mview or in cxcalc the whole process takes minutes and in cxcalc no value is generated due
to a time-out error. The command would be like:
| Code: |
| cxcalc msa "COC(=O)[C@H]1C2[C@]3(C)CC[C@@H](O[Si](C)(C)C)[C@@]2(OC3=O)C4CC[C@@H]5C[C@]14CC5=C" |
Solvent accessible area (3D) of
"COC(=O)[C@H]1C2[C@]3(C)CC[C@@H](O[Si](C)(C)C)[C@@]2(OC3=O)C4CC[C@@H]5C[C@]14CC5=C"
Both started in two independent windows with mview 5.0.1
| Code: |
coming from 3D (strict geometry) (calculation time seconds) ASA = 527.62 ASA+ = 416.20 ASA- = 111.41 ASA_H = 483.34 ASA_P = 44.27 Coming from 2D (calculation time minutes) ASA = 556.03 ASA+ = 462.80 ASA- = 93.23 ASA_H = 509.27 ASA_P = 46.76 |
One problem(?) is that the configuration switched, so in the two cases there are two different molecule values exported,
which is somehow weird.
methyl (1R,5R,8R,9S,11S,14R)-11-methyl-6- methylidene-16-oxo-14-[(trimethylsilyl)oxy]-15- oxapentacyclo[9.3.2.1^{5,8}.0^{1,10}.0^{2,8}]heptadecane-9- carboxylate
methyl (1S,2S,5R,8R,9S,10R,11S,14R)-11-methyl-6- methylidene-16-oxo-14-[(trimethylsilyl)oxy]-15- oxapentacyclo[9.3.2.1^{5,8}.0^{1,10}.0^{2,8}]heptadecane-9- carboxylate
Cheers
Tobias
| Code: |
| cxcalc msa "COC(=O)[C@H]1C2[C@]3(C)CC[C@@H](O[Si](C)(C)C)[C@@]2(OC3=O)C4CC[C@@H]5C[C@]14CC5=C" |