NO binding modes in pdb and csv files generated in the output of Analysis

[sheng19331]

Hi

Thanks for the great tool!
I have tested the script for some of our samples. I think all the remaining parts works fine for me. Only for the output of the binding modes, i could not get neither the summary top_10 csv file in the working folder or the pdb files in the folder of Binding_Modes_Markov_States. Does it have some thresholds for generating such files? Could you help me to figure out the problem? Since those functionality would be important for me to analyse my results. Thanks.

[talagayev]

@sheng19331 Hey,

Yes the default threshold for the interactions that would be taken into account for generating the Binding Modes is 40%. Could you look at the barcodes to see if you have barcodes above 40% and if it is not the case you can reduce the threshold by using the -b option in openmmdl_analysis

so that it would be:
openmmdl_analysis -t {path/to/topology} -d {path/to/trajectory} -n {Ligand_name} -b 30
for a threshold of 30%

Best,
Valerij

[sheng19331]

Valerij

Thanks for your reply.
Yeah, i do consider this possibility and tried to reduce the -b to 30, but still no pdb outputs. For the interaction barcodes, i do have quite a few interactions with more than 40% occupancy during the simulation. I have attached below with the barcodes of three main interactions i have after the anylsis. The thing is i do get a summarised png file with 10 binding modes listed based on your documents. The first one has occurence of 44.7 and the second one is 15.5. But i think your '-b' parameter is to define the interaction occurence instead of this binding mode occurence, right? If i got this binding mode summary png file, i may get the represented pdb file as well based on your script.

Best
Sheng

[talagayev]

@sheng19331 Ah my bad I thought there were no interactions above 40% but now I see,
I discussed it with @NDoering99 and we figured out the reason, he will write here in the comments the command line you need to use.

Best,
Valerij

[NDoering99]

Hi @sheng19331 ,

the top_10_csv and representative frame PDB files will only be written out when adding the "-rep True" flag, as calculating the most representative frame for each b-mode is relatively time-consuming. So to fix your issue, just add "-rep True" to your initial input, and then the files will be generated. I hope this helps (and the documentation will be fixed soon 😄).

Best regards,
Niklas

[sheng19331]

Hi @sheng19331 ,

the top_10_csv and representative frame PDB files will only be written out when adding the "-rep True" flag, as calculating the most representative frame for each b-mode is relatively time-consuming. So to fix your issue, just add "-rep True" to your initial input, and then the files will be generated. I hope this helps (and the documentation will be fixed soon 😄).

Best regards, Niklas

Great! Thanks for figuring out the problem. yeah, as you said. it would be better to add this option in the documentation.