difference in 3d co-ordinates?

Hello again,





And my query for today is...Given the two SMILES codes





a) BrC1=CC(=O)Oc2ccccc12





b) Brc1ccc(cc1)-c2ccc(C=O)o2





if I copy these into MarvinSketch 3.5.5 and complete clean -> 3D -> fine


and then look at source I see following 3D sdf entry:





a)





Marvin 12130507393D





12 13 0 0 0 0 999 V2000


2.4356 -2.1233 -0.0248 Br 0 0 0 0 0 0 0 0 0 0 0 0


1.3341 -0.6560 -0.0077 C 0 0 0 0 0 0 0 0 0 0 0 0


1.9865 0.5180 0.0156 C 0 0 0 0 0 0 0 0 0 0 0 0


1.3013 1.6865 0.0300 C 0 0 0 0 0 0 0 0 0 0 0 0


1.9286 2.7318 0.0789 O 0 0 0 0 0 0 0 0 0 0 0 0


-0.0496 1.7445 -0.0037 O 0 0 0 0 0 0 0 0 0 0 0 0


-0.7554 0.5560 -0.0110 C 0 0 0 0 0 0 0 0 0 0 0 0


-2.1781 0.6341 -0.0127 C 0 0 0 0 0 0 0 0 0 0 0 0


-2.8979 -0.5936 -0.0158 C 0 0 0 0 0 0 0 0 0 0 0 0


-2.2304 -1.8532 -0.0179 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.8098 -1.9325 -0.0172 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.0649 -0.7124 -0.0137 C 0 0 0 0 0 0 0 0 0 0 0 0


1 2 1 0 0 0 0


2 3 2 0 0 0 0


3 4 1 0 0 0 0


4 5 2 0 0 0 0


4 6 1 0 0 0 0


6 7 1 0 0 0 0


7 8 4 0 0 0 0


8 9 4 0 0 0 0


9 10 4 0 0 0 0


10 11 4 0 0 0 0


11 12 4 0 0 0 0


2 12 1 0 0 0 0


7 12 4 0 0 0 0


M END


$$$$





b)





Marvin 12130507353D





14 15 0 0 0 0 999 V2000


2.9367 4.3353 0.6564 Br 0 0 0 0 0 0 0 0 0 0 0 0


2.0233 2.7249 0.5175 C 0 0 0 0 0 0 0 0 0 0 0 0


2.8044 1.5315 0.4665 C 0 0 0 0 0 0 0 0 0 0 0 0


2.0931 0.3111 0.3042 C 0 0 0 0 0 0 0 0 0 0 0 0


0.6561 0.3202 0.1964 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.0844 1.5500 0.2885 C 0 0 0 0 0 0 0 0 0 0 0 0


0.5993 2.7858 0.4509 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.0967 -0.8749 -0.0521 C 0 0 0 0 0 0 0 0 0 0 0 0


0.3383 -2.1982 -0.2298 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.7516 -3.0171 -0.5000 C 0 0 0 0 0 0 0 0 0 0 0 0


-1.8540 -2.1755 -0.4807 C 0 0 0 0 0 0 0 0 0 0 0 0


-3.1388 -2.5771 -0.7129 C 0 0 0 0 0 0 0 0 0 0 0 0


-4.0793 -1.7984 -0.6976 O 0 0 0 0 0 0 0 0 0 0 0 0


-1.4464 -0.9175 -0.2072 O 0 0 0 0 0 0 0 0 0 0 0 0


1 2 1 0 0 0 0


2 3 4 0 0 0 0


3 4 4 0 0 0 0


4 5 4 0 0 0 0


5 6 4 0 0 0 0


6 7 4 0 0 0 0


2 7 4 0 0 0 0


5 8 1 0 0 0 0


8 9 4 0 0 0 0


9 10 4 0 0 0 0


10 11 4 0 0 0 0


11 12 1 0 0 0 0


12 13 2 0 0 0 0


11 14 4 0 0 0 0


8 14 4 0 0 0 0


M END


$$$$





If I process the same two SMILES programmaticaly (using 3.0.12 Jchem.jar) as below,


molecule.clean (3,"S{fine}",null);


I get the following 3D sdf entry





a)





Marvin 12130507203D





12 13 0 0 0 0 999 V2000


-0.5663 3.1989 0.0000 Br 0 0 0 0 0 0 0 0 0 0 0 0


0.1302 1.5012 -0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0


1.4755 1.4300 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0


2.0817 0.2079 -0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0


3.2998 0.1341 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0


1.3953 -0.8489 -0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0


0.1092 -0.9116 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.5379 -2.1814 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0


-1.9617 -2.1824 -0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0


-2.7142 -0.9717 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0


-2.0702 0.2973 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.6413 0.3267 -0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0


1 2 1 0 0 0 0


2 3 2 0 0 0 0


3 4 1 0 0 0 0


4 5 2 0 0 0 0


4 6 1 0 0 0 0


6 7 1 0 0 0 0


7 8 4 0 0 0 0


8 9 4 0 0 0 0


9 10 4 0 0 0 0


10 11 4 0 0 0 0


11 12 4 0 0 0 0


2 12 1 0 0 0 0


7 12 4 0 0 0 0


M END


$$$$





b)





Marvin 12130507263D





14 15 0 0 0 0 999 V2000


2.3124 4.6925 0.6279 Br 0 0 0 0 0 0 0 0 0 0 0 0


1.3598 3.1022 0.5260 C 0 0 0 0 0 0 0 0 0 0 0 0


2.1171 1.8945 0.4632 C 0 0 0 0 0 0 0 0 0 0 0 0


1.3802 0.6884 0.3127 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.0552 0.7239 0.2272 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.7745 1.9670 0.3408 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.0639 3.1900 0.4916 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.7191 -0.5199 -0.0351 C 0 0 0 0 0 0 0 0 0 0 0 0


-2.0812 -0.7980 -0.2304 C 0 0 0 0 0 0 0 0 0 0 0 0


-2.2449 -2.1467 -0.5254 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.9663 -2.6839 -0.5022 C 0 0 0 0 0 0 0 0 0 0 0 0


-0.6647 -3.9915 -0.7564 C 0 0 0 0 0 0 0 0 0 0 0 0


0.4824 -4.4096 -0.7387 O 0 0 0 0 0 0 0 0 0 0 0 0


-0.0821 -1.7088 -0.2012 O 0 0 0 0 0 0 0 0 0 0 0 0


1 2 1 0 0 0 0


2 3 4 0 0 0 0


3 4 4 0 0 0 0


4 5 4 0 0 0 0


5 6 4 0 0 0 0


6 7 4 0 0 0 0


2 7 4 0 0 0 0


5 8 1 0 0 0 0


8 9 4 0 0 0 0


9 10 4 0 0 0 0


10 11 4 0 0 0 0


11 12 1 0 0 0 0


12 13 2 0 0 0 0


11 14 4 0 0 0 0


8 14 4 0 0 0 0


M END


$$$$





As you can see the two sets of data are different.


Why is this - different implementation of 3d co-ordinate generation between the two? Please explain, I would expect them to be identical, other than if they are different versions.





Is this a Marvin version number?


Why does this occur 3 times : 12130507393D


And this only once: 12130507263D





???





Many thanks for your comments,


Daniel.

inhibox wrote:

Hello again, And my query for today is...Given the two SMILES codes a) BrC1=CC(=O)Oc2ccccc12 b) Brc1ccc(cc1)-c2ccc(C=O)o2 if I copy these into MarvinSketch 3.5.5 and complete clean -> 3D -> fine and then look at source I see following 3D sdf entry: ... If I process the same two SMILES programmaticaly (using 3.0.12 Jchem.jar) as below, molecule.clean (3,"S{fine}",null); I get the following 3D sdf entry ... As you can see the two sets of data are different. Why is this - different implementation of 3d co-ordinate generation between the two? Please explain, I would expect them to be identical, other than if they are different versions. Is this a Marvin version number? Why does this occur 3 times : 12130507393D And this only once: 12130507263D ??? Many thanks for your comments, Daniel.

Hi Daniel,





Processing Clean3D from MSketch may result in different coordinates due to the Clean2D which is peformed in MSketch. The 2D molecule may have additional information (like cis/trans) which was created by Clean3D but such information is not present in the SMILES string. I suggest using the "programmed" way or molconvert for best results.





Ödön

inhibox wrote:

Is this a Marvin version number? Why does this occur 3 times : 12130507393D And this only once: 12130507263D

Actually the last part of that script specifies the dimension (3D), and the beginning refers to the date (mmddyy) and time (hhmm) when the file was saved. So "12130507393D" means that it is a 3D file which was saved at 7:39 on the 13-th of December, 2005.





Best regards,


Szabolcs

inhibox wrote:

... As you can see the two sets of data are different. Why is this - different implementation of 3d co-ordinate generation between the two? Please explain, I would expect them to be identical, other than if they are different versions. ... Many thanks for your comments, Daniel.

Hi Daniel,





We examined your case and it turned out that while what I described in my previous answer may cause slight differences in coordinates or even may result in a different conformer, in your case, the different results are due to the use of different Marvin releases. There was a change in Clean3D after Marvin 3.5.5 and some other changes may affect the results of the Clean3D process in the forthcoming Marvin release. If you upgrade your Marvin version now you should get identical results to what you reported with your current JChem version.





I hope this helped,





Ödön