How to interpret lddtfull?

[mwittep]

Context

I am comparing the structure of proteins, but rather than focusing on the general similarity, I want to focus on the similarity at specific domains (for example, around known binding sites). I know the position of these binding sites on the sequence, and I would like to transfer this knowledge to the structural alignment.

I thought that using lddtfull would provide such information, but I am confused that it does not have the same length as the alignment.

I know that foldseek reduces the structure to the 3Di alphabet, but is there any way of still transfering this kind of information?

Expected Behavior

lddtfull should have the same length as the alignment.

Current Behavior

The length of the sequence alignment and the structural alignment differ.

Your Environment

Include as many relevant details about the environment you experienced the bug in.

• Git commit used: 7da7d09
• Which foldseek version was used: statically-compiled