`add-elem-info` (repair PDB element columns)¶

add-elem-info fills in or corrects the element symbols (columns 77–78) of the ATOM and HETATM records in a PDB file. It infers each element from the atom name and the residue name, keeps element fields that already hold a valid symbol, and changes nothing else in the file.

What it is for¶

PDB files without element columns: structures from modeling tools or molecular dynamics (MD) that leave columns 77–78 blank.
Wrong element symbols: re-infer every column with --overwrite-elem.
Preparing input for other commands: all repairs blank element fields by itself with the same rules. Run add-elem-info before a standalone command such as extract, or when a symbol is wrong rather than blank.

Examples¶

1. Write `<input>_add_elem.pdb`¶

Fill the element columns and write the result next to the input.

mlmm add-elem-info -i 1abc.pdb

The console prints [add-elem-info] Wrote: 1abc_add_elem.pdb and the counts total atoms, assigned/updated, and kept existing. With no WARNING line, every atom has an element.

2. Choose the output file¶

mlmm add-elem-info -i 1abc.pdb -o 1abc_fixed.pdb

3. Overwrite the input¶

Replace the input file itself.

mlmm add-elem-info -i 1abc.pdb --overwrite

How it works¶

Reading the records: add-elem-info reads every line, follows MODEL blocks, and looks only at ATOM and HETATM records.
Keeping valid symbols: An element field that already holds a valid symbol (or EP for a water virtual site) is kept. Blank or unrecognized fields are repaired; --overwrite-elem re-infers every field.
Inferring the element: The element comes from the four-character atom name (columns 13–16) and the residue name. Ion residues give the element of the ion; amino acids, nucleic acids, and water follow the usual naming (H/D, C, N, O, P, S, Se); other ligands use the column alignment of the atom name (NA is N, NA is Na), including LEaP halogens such as CL1 and hydrogen names such as HG11.
Writing and the summary: Only columns 77–78 of the repaired records change. The console reports the totals, the counts per element, and the atoms that could not be assigned.

Output files¶

The repaired PDB: <input>_add_elem.pdb by default, the path given with -o, or the input file itself with --overwrite and no -o.
Console summary: total atoms, assigned/updated (atoms whose element column changed), kept existing, and assignment breakdown (counts per element). Atoms that could not be assigned are left unchanged and listed after [add-elem-info] WARNING: Could not confidently assign N atoms; left unchanged., up to 50 of them. Type the element symbol of each listed atom into columns 77–78 by hand, right-aligned; --overwrite-elem uses the same rules and will not assign them.

Main options¶

Option	Type	Default	Description
`-i, --input`	path	(required)	Input PDB file
`-o, --output`	path	`None`	Output PDB file; without it, `<input>_add_elem.pdb`
`--overwrite/--no-overwrite`	flag	`False`	Overwrite the input file when `-o` is not given
`--overwrite-elem/--no-overwrite-elem`	flag	`False`	Re-infer element columns that already hold a valid symbol

See the generated CLI reference for every option.

Notes¶

What changes: Every input line is preserved byte-for-byte except columns 77–78 of the ATOM and HETATM records being repaired. HEADER, REMARK, CONECT, ANISOU, and the charge columns (79–80) are written unchanged.
Models and chains: ATOM and HETATM records in every model, chain, and residue are processed.
Special names: deuterium labels become H, selenium (SE*) is recognized as Se, and halogens are recognized automatically.
Two different flags: --overwrite-elem decides which element columns are re-inferred; --overwrite decides only where the file is written.
Writing to the input path: when -o names the input file, including through a symbolic link, --overwrite is required; without it the run stops with an error.
Exit codes: see Exit codes.

add-elem-info (repair PDB element columns)¶