Printing out ATOM IDS corresponding each argument while rendering using plumed writer

[sbhakat]

Hi,

I managed to get the TICA coefficients in a Plumed format for my contact features (see the attached document). However I want a smart way print out the ATOM IDs associated with the contact features so that I can construct the plumed file easily. Suggestions will be highly appreciated.

Please see the bottom of the attached document for the code

bpti_msm_met52n.pdf

[msultan]

hmm, thats an msmbuilder issue i think. Can you try the following :

feat = ContactFeaturizer(scheme='ca', contacts=np.array(contactlist))

instead of what you have now.

feat = ContactFeaturizer(scheme='CA',contacts=np.array(contactlist))

i think the capitalized "CA" is causing some weird edge case here.

[sbhakat]

I managed to get it working by following script

atind=[5,37,43,55,75,85,104,127,141,147,168,182,197,203,213,235,245,269,288,307,331,352,372,393,407,417,439,449,456,475,485,502,516,536,552,573,580,587,597,621,631,653,677,691,705,725,747,758,768,783,795,805,822,846,860,870,877,884]
for f in feat.describe_features(ds[0]):
    [a,b]=f['resids']
    print("DISTANCE ATOMS=%d,%d"%(atind[a-1], atind[b-1]))

Now another silly problem though, I made the plumed.dat file however while testing with driver it is giving an error like ERROR in input to action COMBINE with label tic0 : keyword PERIODIC is compulsory for this action

But I already set the PERIODIC=NO I am not sure why this error. Any input? File attached herewith.

plumed.txt

[msultan]

This must be that the distances that need to have "PERIODIC=FALSE" not the tica coordinate.

Also, you could try to just update the featurizer since it can automatically generate the correct data frame for you. That way their is one less source of error for if you want to try more/less contacts. If you end up using this, make sure to compare the plumed output with the tica results.