Constrain or skip of opt in aISS workflow.

[lizhiq16]

I have some "flexible" molecules obtained from crystal structures or DFT calculations, and I want to use aISS to generate aggregate structures for them. When directly using the default aISS process, I find that the interactions between the two molecules are overemphasized, resulting in closely bound structures with significant distortions in molecular conformations. Therefore, I would like to know:

1. Is it possible to perform only a procedure similar to rigid docking through commands line option, while ignoring the subsequent structure optimization steps?
2. The aISS manual mentions that constraining and fixing of atoms is possible, but it does not provide any examples. Since the input contains two structure files, could you please guide me on how to construct the constraining and fixing part of the file to restrict the structure of one molecule or individually constrain the conformations of both molecules?

By the way, when I perform an additional pocket search, I obtain a conformation where the two molecules are fused together ( not two indivdual molecules anymore). Is there a bug here?

[cplett]

Hi, the constraining and fixing work similar to xtb which is explained here. You have to provide an input file and then specify it with the command line call via --input <file>. The numbers that are provided as atom specifiers in this file start to count with the molecule you provide first. If your first molecule contains 3 atoms, a 4 in the input file would be related to the first atom of your second input molecule. However, you could also take the structures right before the optimization as either best_after_gen.xyz, or structures_after_gen.xyz depending on the xtb version you are using.

The pocket search does sometimes have some problems which is the reason why it is not the default.

[manuelperezesc]

Dear @cplett,

I am using the precompiled xtb 6.7.0 version and for some reason I do not get the structures_after_gen.xyz file. I thought it had something to do with the use of the pocket search, but it does not work when I remove the keyword.

I just got best_after_gen.xyz, best.xyz, final_structures.xyz (which I guess it's the final result, after optimization) and optimized_structures.xyz, which seems to be the same. Am I missing something?

[camattelaer]

+1 I would also be interested in skipping the optimization of the structure, i.e. stopping the protocol after the generation without having to kill the process.

Kind regards

Charles-Alexandre