search

SPICA-group / spica-tools

Tools for coarse-grained molecular dynamics simulations using the SPICA force field
https://spica-group.github.io/spica-tools/
MIT License
11 stars 4 forks source link

The 'ENM' command encountered an "IndexError: list index out of range" error #32

Open xiaoyu2260 opened 3 weeks ago

xiaoyu2260 commented 3 weeks ago

Thank you very much for your reply. I have a question regarding the ENM command.

1.I am currently using the PDB ID: 6DS5. 2.For a single protein structure, the issue I described previously can be resolved. However, if a single protein contains two alpha-helix structures, additional bond information is still generated between the two helices. For instance, this occurs with Chain A of 6DS5. (Please see Fig. 1 for reference.) 6ds5.zip fig 1

yskmiyazaki commented 3 weeks ago

@xiaoyu2260 Thank you for sharing the pdb file. The elastic networks in the output you got through the ENM command are correct at least in the SPICA FF. Applying elastic networks depends solely on distances between coarse-grained backbone segments and SPICA has adopted this protocol for ENM. The Martini FF seems to employ the same way for ENM. See https://cgmartini.nl/docs/tutorials/Martini3/ProteinsI/.

The channel protein consisting of the multiple domains may not maintain its structure during SPICA-CG MD simulations without the elastic network bonds between the two helices. The association free energy of the improved SPICA protein model shows good agreement with experiments, but maintaining the secondary and tertiary structure of protein found in experiments is still quite difficult in CG-MD.

Fig S8 in the SI of the first SPICA protein model paper https://pubs.acs.org/doi/suppl/10.1021/acs.jctc.1c01207/suppl_file/ct1c01207_si_001.pdf shows snapshots of membrane protein complex, and the elastic networks were applied between intramolecules there, meaning elastic network bonds between the helices were generated there ( not between intermolecules). This simulation with SPICA is to evaluate the stability of the protein complex in the membrane, assuming the secondary/tertiary structures of each domain were stable. So, the SPICA protein model with ENM was okay.

Using this kind of protein models based on ENM or other protein models established without ENM depends on the objective of your research. If you are going to see the structure change of membrane protein during MD simulation, it is better to consider other CG force field, Go-Martini, SIRAH, and so on.