Dssp ValueError: could not broadcast input array from shape (71,) into shape (70,)

[intomybioverse]

I am trying to run the biotite secondary structure analysis program in wsl to study the secondary structure of a protein after md simulation. The website link for the python code is provided below.

[Secondary structure of protein during an MD simulation using Biotite] (Secondary structure during an MD simulation — Biotite 0.41.0 documentation )

The error message goes like this,
sse[idx] = app.get_sse()
ValueError: could not broadcast input array from shape (71,) into shape (70,)

It is working for the lysozyme.xtc (given as example) and other md trajectory files that I have used, but not this one. How do I solve this issue?

[padix-key]

Probably this has something to do with residue recognition. Can you provide the template structure file?

[intomybioverse]

template.txt
I tried with the .gro file (produced after equilibriation) and now the error is as follows.

(biotite_env) (base) sobs@BEE:~/all_my_data/SCR_wtn3m$ python3 trajectory_sse.py
Traceback (most recent call last):
File "/home/sobs/all_my_data/SCR_wtn3m/trajectory_ss.py", line 49, in
app.join()
File "/home/sobs/.local/biotite_env/lib/python3.10/site-packages/biotite/application/application.py", line 63, in wrapper
return func(*args, **kwargs)
File "/home/sobs/.local/biotite_env/lib/python3.10/site-packages/biotite/application/localapp.py", line 257, in join
self.evaluate()
File "/home/sobs/.local/biotite_env/lib/python3.10/site-packages/biotite/application/dssp/app.py", line 90, in evaluate
super().evaluate()
File "/home/sobs/.local/biotite_env/lib/python3.10/site-packages/biotite/application/localapp.py", line 278, in evaluate
raise SubprocessError(
subprocess.SubprocessError: 'mkdssp' returned with exit code 1: DSSP could not be created due to an error: invalid formatted floating point number '-7.6293945e-06'

Dssp version 2.3.0 is installed but it is not responding. Then I uninstalled and reinstalled it from this site "https://github.com/cmbi/dssp/archive/refs/tags/2.3.0.tar.gz", installed it and still it is showing mkdssp version 3.1.4 .

.py file: https://www.biotite-python.org/examples/gallery/structure/trajectory_sse.html
I have attached the template file too.

[padix-key]

Sorry for the later answer, I was a few days on vacation. The problem is atom naming: The trajectory from the example gallery was created with GROMACS, whose atom naming sometimes deviate from the standard PDB nomenclature. Specifically, the C-terminal O and OXT are named O1 and O2. However without the O DSSP does not recognize the terminal residue as amino acid, and skips it. Hence, the array with the secondary structure elements has one element less in the example:

# DSSP does not assign an SSE to the last residue -> -1
sse = np.empty((traj.shape[0], struc.get_residue_count(traj)-1), dtype='U1')

However, in your template the atoms are named correctly. Hence DSSP returns has the correct length. Since the sse array is still shortened by 1, this gives an exception due to the mismatch.

To fix this you can rename the problematic C-terminal atoms correctly (i.e. O1 -> O in the last residue), in case it not already is. Then you can remove the -1 in the sse creation.