`user_custom`¶

Use this attribute to bring your own CHARMM-format topology, parameter, or stream files (e.g. a CGenFF stream for a small-molecule ligand) into a pestifer build, so that psfgen and NAMD pick them up without any further wiring.

How it works¶

Put your file(s) in one or more directories on disk.
List each directory under searchpath. Every CHARMM-format file (.rtf, .top, .str, .prm) in those directories is registered with pestifer and indexed by the RESI / PRES records it contains.
When psfgen builds the system, any residue name in your input PDB whose definition lives in one of these files is automatically resolved: the file is copied into the run directory, a topology <file> line is emitted, and the file is registered in the pipeline so NAMD’s parameter file list includes it as well.
searchpath entries may contain ~/ (the user’s home directory) or shell environment-variable references such as $HOME/foo or ${SCRATCH}/charmm; both are expanded before pestifer probes the directory.

Per-user toppar cache¶

~/.pestifer/toppar/ is a well-known per-user cache directory for custom CHARMM stream files. When it exists, pestifer automatically prepends it to searchpath so any .str / .rtf / .prm files dropped there are picked up without requiring a user_custom block in the YAML. This is the recommended destination for the .str files generated by make-ligand-mol2 (after running them through the CGenFF binary or web tool): once a parameter file for, e.g., AP5 lives in ~/.pestifer/toppar/, every subsequent build that references that residue resolves it transparently.

Any directory the user lists in searchpath is loaded after the cache, so a project-local entry overrides anything in ~/.pestifer/toppar/ (last-loaded wins). When this happens pestifer emits a warning naming both files.

Segtype classification¶

Every RESI / PRES found in a user_custom file is also registered with pestifer’s segtype machinery so the molecule parser can place those residues into segments. By default each new residue is classified as ligand. Use the segtypes knob below to override (e.g. mark a residue as cofactor instead). An explicit psfgen.segtypes entry for the same resname always wins.

Caveats¶

searchpath entries are directories; the per-extension lists below (rtf, str, prm) take basenames only – any directory separator in those entries is rejected.
A single .str file that contains both read rtf … END and read para … END blocks is sufficient – you do not need to list it under more than one of rtf / prm.
custom is for files that ship with pestifer itself; user-supplied files belong here under user_custom.
Atom names in your input PDB must match the atom names declared in the RESI block of the stream file. If they don’t (e.g. RCSB PDBs typically use a different naming convention than CGenFF-generated streams), psfgen will build the residue topology but leave each unmatched atom at (0, 0, 0). Add psfgen.aliases.atom entries to bridge the two naming schemes.

Example:¶

charmmff:
  user_custom:
    searchpath:
    - /path/to/my/custom/search/path/
    rtf:
    - a_file.rtf
    str:
    - a_file.str
    prm:
    - a_file.prm

Container-like attributes:

user_custom¶