pymol-scripts

Three small PyMOL scripts for preparing protein structures for design and interface analysis. Each is standalone — drop it next to your structure and run.

renumber_and_merge_chains.py

Fetches a PDB entry, renumbers all protein chains into one continuous residue series, and merges them into a single chain.

run renumber_and_merge_chains.py
renumber_and_mergeChains 1crn        # merge into chain A (default)
renumber_and_mergeChains 1crn, B     # merge into chain B

Why this matters: many design and folding tools want a single chain with monotonic numbering, but deposited structures arrive as multiple chains with overlapping or gapped numbering. This flattens them. Non-protein content (DNA, RNA, ligands) is left alone — the chain selection is restricted to polymer.protein.

Insertion codes are sidestepped by reading residue numbers off name CA and skipping any resi that isn't purely numeric.

renumber_and_merge_chains_with_gap.py

Same operation, but leaves a 3-residue numbering gap between consecutive chains.

run renumber_and_merge_chains_with_gap.py
renumber_and_mergeChainsGap 1crn

Use this variant when the merged chain feeds a structure-prediction or design tool that infers chain breaks from numbering discontinuity. A flat continuous series (the other script) implies the chains are covalently connected, which will make a folding model try to bridge termini that are actually separate. The gap preserves the break.

hotspot_selection.py

Given a set of hotspot residues, finds every residue outside a distance shell around them and prints the list sorted by chain and residue number.

pymol -cq hotspot_selection.py

Edit the top of main() to set your inputs:

pdb_file = "myprotein.pdb"
hotspot_res_list = ["A/10", "A/25", "A/100"]   # chain/resi
threshold = 5.0                                 # Angstroms

The selection uses byres, so whole residues are returned rather than stray atoms. Handy for building the "everything I'm allowed to redesign" set — the complement of the interface — e.g. as a fixed/mutable residue mask for ProteinMPNN or an RFdiffusion contig.

There's a commented-out block at the end that renumbers the selected residues consecutively via cmd.alter. It's off by default because the script is usually wanted for the residue list alone; uncomment it if you want the structure modified in place.

Requirements

PyMOL with Python scripting — open-source build is fine:

conda install -c conda-forge pymol-open-source

The two renumber_and_merge_* scripts call cmd.fetch, so they need network access to pull from RCSB. Point them at a local file with cmd.load instead if you're working offline.

License

MIT — see LICENSE.