Select conformers documentation?

Do you have any documentation on how "Select conformers" works? Does it show all local minima, or local minima below a certain energy barrier, or what? Is there a way to control how many conformers are generated? Etc.

The select conformer option is designed for getting the conformers for smaller molecules. The interactive (through MSketch or MView) version has no capabilities to accept extra options. The current limits are: max. 20 conformers if the molecule contain chains or max. 50 conformers if the molecule contains only ring(s).





The conformer search can also be used in batch mode through molconvert. See the extra options executing "molconvert -h" or for the 3D clean detailed options "molconvert -H3D".





The following lines has been taken from the output of "molconvert -H3D":





[ca] Conformational analysis (case insensitive)


Fill a specific molecule property with conformers found.


For compatibility reasons, [ca] option will invoke Cleaner strategy 5.





...





[writeconfs] Write conformers to mol file (case insensitive)


{ <FILENAME PREFIX> }


Write all generated conformers for each fragment into molfiles.


Files generated (in working directory): <FILENAME PREFIX>.<FRAGMENT>.mol


default value: conformers


...





[CPfargmulti] Default fragment multiplicity (case sensitive)


[fm] (case insensitive)


Note that setting cleaner strategy will overwrite these settings.


{ <integer value for OPT_CP_FRAGMULTI_CHAIN> }


current value: 20


{ <integer value for OPT_CP_FRAGMULTI_RING> }


current value: 50


{ <integer value for OPT_CP_FRAGMULTI_RINGBEFORECLOSE> }


current value: 50








For conformer search in batch mode try something like the following example:





# This will produce no more than 100 conformers in the file "filename_for_conformers.0.mol" in MDL mol file format


# If the molecule has more than one fragment then the conformers for the fragments will be written to separate files.





molconvert -3:S0[ca][fm]{100}{100}{100}[writeconfs]{filename_for_conformers} mol <input-file>





When using large numbers (>500) for conformer limit changing the memory limit for java might be necessary by editing the "molconvert" script file.





The conformer search found to be very useful for finding conformers of partially flexible ring systems.





If the energy difference of the conformers is a concern, please keep in mind that the DREIDING force field, which is used for calculating the energy of the conformers, has some limitations. In such cases the user is adviced to use the results of the conformer search as input for some higher level, e.g. quantum chemical studies to calculate more accurate energies.

Thanks for the answer, but you didn't really answer my question. First, please know that I am NOT a cheminformaticist. I am merely an organic chemist and a MarvinSketch user. I am not capable of converting batches, executing scripts, or the like.





I drew butane in its highest-energy conformation (0 degree dihedral angle) in MarvinSketch. I cleaned it in 3D. I then chose Select conformers. A window opened with many, many pictures of the anti conformer of butane. Neither the gauche nor any of the eclipsed conformers were present.





So, what are the parameters of your algorithm that it can't even find the gauche conformer of butane? And why does Marvin show multiple pictures of the same conformer?

Sorry for the too technical advices. I assume that you also added the Hydrogens explicitly to the butane. This shows a known problem in the current released version: it handles each Hydrogene atom as a different one and provides all possible permutations of the Hydrogens in similar position as different conformers. Since the number of conformers is limited to 20 for chains by default, the result was variations for the anti conformer.





My advice for the moment is adding hydrogens after doing the conformational search.





Sorry for the technical details but still, I give an axample how to get the first 100 conformers of n-nonane (the conformers come in the file named nonane_confs.0.mol):





molconvert mol -3:S0[ca][fm]{100}{100}{100}[writeconfs]{nonane_confs} -s "CCCCCCCCC"





Then the following command will hydrogenize and then optimize the conformers. The final energies are written to the SDF file.





molconvert sdf -3:c01[hydrogenize]o102E nonane_confs.0.mol >nonane_confs_H.sdf





Sorry for the inconveniences, this is a new and improving feature of Marvin.

ofarkas wrote:

This shows a known problem in the current released version: it handles each Hydrogen atom as a different one and provides all possible permutations of the Hydrogens in similar position as different conformers. Since the number of conformers is limited to 20 for chains by default, the result was variations for the anti conformer. Sorry for the inconveniences, this is a new and improving feature of Marvin.

Ah. No wonder. OK, I look forward to using it when it is working better, i.e., treats the three rotamers of a CH3 group as identical.

I tried to use the command:


molconvert mol -3:S0[ca][fm]{100}{100}{100}[writeconfs]{nonane_confs} -s "CCCCCCCCC"


but I got the error message:





molconvert: No match





then I gave the command:





molconvert mol -3:S0\[ca\]\[fm\]{100}{100}{100}\[writeconfs\]{nonane_confs} -s "CCCCCCCCC"





then I got the error message:


Non recognized option character: _


Unknown Clean3D option: n


Unknown Clean3D option: n


Unknown Clean3D option: a


Unknown Clean3D option: n


Unknown Clean3D option: n


Unknown Clean3D option: f


Unknown Clean3D option: s





and one file "conformers.0.mol" is created which contains one conformation.


Could someone please tell me what is going on? Thanks.

Sorry for the late answer, I did not check this topic for new questions.


Please, as it is normally required in the forum, let me know about your operating system, java and Marvin version. Please answer with "quote" to make sure I get an e-mail concerning your response.





Best wishes,





Ödön

wangsantacruz wrote:

I tried to use the command: molconvert mol -3:S0[ca][fm]{100}{100}{100}[writeconfs]{nonane_confs} -s "CCCCCCCCC" but I got the error message: molconvert: No match then I gave the command: molconvert mol -3:S0\[ca\]\[fm\]{100}{100}{100}\[writeconfs\]{nonane_confs} -s "CCCCCCCCC" then I got the error message: Non recognized option character: _ Unknown Clean3D option: n Unknown Clean3D option: n Unknown Clean3D option: a Unknown Clean3D option: n Unknown Clean3D option: n Unknown Clean3D option: f Unknown Clean3D option: s and one file "conformers.0.mol" is created which contains one conformation. Could someone please tell me what is going on? Thanks.