Jungfraujoch/image_analysis/structure_refinement/monomers/update_monomers.py

#!/usr/bin/env python3
# SPDX-FileCopyrightText: 2026 Filip Leonarski, Paul Scherrer Institute <filip.leonarski@psi.ch>
# SPDX-License-Identifier: GPL-3.0-only
"""Copy the subset of the CCP4 Monomer Library (https://github.com/MonomerLibrary/monomers, LGPL-3.0)
that rugnux's restrained refinement carries in its binary into this directory.

    update_monomers.py /path/to/monomers

COMPOUNDS below is THE list: a compound named there is restrained by the refinement, everything else
is held fixed. To add one, add its code and run this again; CMake embeds every .cif of this directory
(MonomerLibrary.cpp), so nothing else needs to change.

The compound files are copied verbatim. links_and_mods.cif is cut down to the links and modifications
that can apply to these compounds (the polymer links, the termini, the disulfide, ...), and ener_lib.cif
to its atom-type table (_lib_atom: the van der Waals and ion radii and the hydrogen-bond types).
metals.json, the library's statistics of metal-ligand distances, becomes metals.cif: for each metal ion of
COMPOUNDS and each donor element of METAL_LIGANDS, the distance per coordination number (the restraints of
metal sites the links of links_and_mods.cif do not describe).
Requires the gemmi Python module."""

import json
import os
import shutil
import sys

import gemmi

AMINO_ACIDS = ["ALA", "ARG", "ASN", "ASP", "CYS", "GLN", "GLU", "GLY", "HIS", "ILE",
               "LEU", "LYS", "MET", "PHE", "PRO", "SER", "THR", "TRP", "TYR", "VAL", "MSE"]
NUCLEOTIDES = ["DA", "DC", "DG", "DT", "A", "C", "G", "U"]
# The modified nucleotides of tRNA and rRNA most often met, polymer residues joined by the same backbone link.
MODIFIED_NUCLEOTIDES = ["PSU", "5MC", "1MA", "OMG", "OMC", "OMU", "H2U", "M2G", "7MG", "2MG", "5MU"]
NUCLEOTIDE_LIGANDS = ["ATP", "ADP", "AMP", "GTP", "GDP", "GMP"]
# The cofactors most often met, and the oxygen a heme iron binds. (Not carbon monoxide: CMO.cif has C-O at 1.25 A,
# where bound CO is near 1.15 A.)
COFACTORS = ["NAD", "NAI", "NAP", "NDP", "FAD", "FMN", "HEM", "HEC", "SAM", "SAH", "OXY"]
ADDITIVES = ["DMS", "PEG", "PGE", "PG4", "1PE", "GOL", "EDO", "SO4", "PO4", "ACT", "MPD"]
WATER_AND_IONS = ["HOH", "NA", "K", "MG", "CA", "ZN", "MN", "FE", "FE2", "NI", "CU", "CO", "CD",
                  "CL", "BR", "IOD"]
COMPOUNDS = (AMINO_ACIDS + NUCLEOTIDES + MODIFIED_NUCLEOTIDES + NUCLEOTIDE_LIGANDS + COFACTORS + ADDITIVES +
             WATER_AND_IONS)
# The donor elements of metal sites whose distances metals.cif carries.
METAL_LIGANDS = ["O", "N", "S"]

# Groups of the polymers above; links and modifications of other groups (sugars) are left out.
GROUPS = {".", "peptide", "P-peptide", "M-peptide", "DNA/RNA", "NON-POLYMER"}


def compound_path(lib, code):
    return os.path.join(lib, code[0].lower(), code + ".cif")


def applies(comp, group):
    return (comp == "." or comp in COMPOUNDS) and group in GROUPS


def trim_links_and_mods(lib, out):
    doc = gemmi.cif.read(os.path.join(lib, "links_and_mods.cif"))
    links = doc.find_block("link_list").find("_chem_link.", ["id", "comp_id_1", "mod_id_1", "group_comp_1",
                                                             "comp_id_2", "mod_id_2", "group_comp_2", "name"])
    mods = doc.find_block("mod_list").find("_chem_mod.", ["id", "name", "comp_id", "group_id"])
    keep_links = [list(r) for r in links if applies(r[1], r[3]) and applies(r[4], r[6])]
    keep_mods = [list(r) for r in mods if applies(r[2], r[3])]
    keep = {"link_" + r[0] for r in keep_links} | {"mod_" + r[0] for r in keep_mods}

    new = gemmi.cif.Document()
    for name, tag, rows in (("link_list", "_chem_link.", keep_links), ("mod_list", "_chem_mod.", keep_mods)):
        block = new.add_new_block(name)
        old = doc.find_block(name).find_loop_item(tag + "id").loop
        loop = block.init_loop(tag, [t[len(tag):] for t in old.tags])
        for r in rows:
            loop.add_row(r)
    for block in doc:
        if block.name in keep:
            new.add_copied_block(block)
    new.write_file(out)
    return len(keep_links), len(keep_mods)


def trim_ener_lib(lib, out):
    doc = gemmi.cif.read(os.path.join(lib, "ener_lib.cif"))
    atoms = doc[0].find_loop_item("_lib_atom.type").loop
    new = gemmi.cif.Document()
    block = new.add_new_block(doc[0].name)
    loop = block.init_loop("_lib_atom.", [t[len("_lib_atom."):] for t in atoms.tags])
    for i in range(atoms.length()):
        loop.add_row([atoms[i, j] for j in range(atoms.width())])
    new.write_file(out)


def trim_metals(lib, out):
    """metals.json -> metals.cif: one row per metal ion of COMPOUNDS, ligand element and coordination number,
    with the mean distance, its standard deviation and the number of sites behind it."""
    with open(os.path.join(lib, "metals.json")) as f:
        table = json.load(f)["metal_coordination"]
    metals = set()
    for code in WATER_AND_IONS:
        for atom in gemmi.make_chemcomp_from_block(gemmi.cif.read(compound_path(lib, code))[-1]).atoms:
            if atom.el.is_metal:
                metals.add(atom.el.name)
    new = gemmi.cif.Document()
    block = new.add_new_block("metal_coordination")
    loop = block.init_loop("_metal_coordination.", ["metal", "ligand", "coordination", "distance", "std", "count"])
    n = 0
    for metal in sorted(metals):
        for ligand in METAL_LIGANDS:
            for row in table.get(metal, {}).get(ligand, []):
                loop.add_row([metal, ligand, str(row["coord"]), "%.4f" % row["mean"], "%.4f" % row["std"],
                              str(row["count"])])
                n += 1
    new.write_file(out)
    return n


def main():
    if len(sys.argv) != 2:
        sys.exit(__doc__)
    lib = sys.argv[1]
    here = os.path.dirname(os.path.abspath(__file__))
    for f in os.listdir(here):
        if f.endswith(".cif"):
            os.remove(os.path.join(here, f))
    for code in COMPOUNDS:
        shutil.copyfile(compound_path(lib, code), os.path.join(here, code + ".cif"))
    n_links, n_mods = trim_links_and_mods(lib, os.path.join(here, "links_and_mods.cif"))
    trim_ener_lib(lib, os.path.join(here, "ener_lib.cif"))
    n_metal_rows = trim_metals(lib, os.path.join(here, "metals.cif"))
    shutil.copyfile(os.path.join(lib, "COPYING"), os.path.join(here, "COPYING"))
    print(f"{len(COMPOUNDS)} compounds, {n_links} links, {n_mods} modifications, {n_metal_rows} metal distances")


if __name__ == "__main__":
    main()