Problem with selecting chains based on PDB for XTC generated from TOP + CRD files

[HoaSoup]

Hello,

I would like to analyze hydrogen bonds from some trajectories, using the original PDB that was used to generate Amber CRD and TOP files. The bonds are between a protein and a peptide. In the original input PDB file, they are of two different chains. I prefer PDB because it seems that TOP files don't store chain IDs.

I also notice that PDB also issues an index number to any TER records it has. I am worried that when the PDB file is translated into TOP, TER records are ignored, particles take the atom indexes of TER and hence the index numbering scheme of TOP will be different from PDB. The consequence is that when I want to select the chains in the PDB files corresponding to the protein and the peptide, the selection would not follow the index scheme of XTC files, which are derived from the scheme of the TOP files.

Is there a workaround for this?

Best regards,
Hoa

[orbeckst]

What is the specific MDAnalysis question? Do you have some code to show? Or is this a general question?