Architecture: mlmm-toolkit¶

1. Overview¶

mlmm-toolkit is a Python CLI that performs ML/MM (ONIOM) enzymatic reaction-path analysis on a complete protein environment. ML/MM here means a hybrid model in which a small reaction core is treated by a machine-learning interatomic potential (ML) and the surrounding protein by a molecular-mechanics (MM) force field, combined through the subtractive ONIOM (Our own N-layered Integrated molecular Orbital and molecular Mechanics) energy scheme.

The all workflow can generate a parm7 topology and encode the ONIOM region split (ML / Movable-MM / Frozen) from structure and region inputs. Multi-input runs search a path; single-input runs require either --scan-lists or --tsopt.

Depending on its mode and flags, all composes stages from extract, mm-parm, MEP search, TS optimization, IRC, frequency analysis, and single-point DFT. TS, thermochemistry, and DFT stages are optional.

The package is laid out as 6 physical layer directories (cli/, workflows/, domain/, backends/, io/, core/). The role and dependency direction of each are summarized in the §2.1 layer table below.

External code imports directly from the layer directory (from mlmm.backends.mlmm_calc import MLMMCore, from mlmm.core.utils import …, import mlmm.io.trj2fig, etc.). §2.4 details the two supported import surfaces.

Three bundled forks (pysisyphus/, thermoanalysis/, hessian_ff/) live at the repo top as repo-internal modules. They are deliberately not the upstream PyPI distributions (and hessian_ff/ has no upstream at all — bundling is mandatory). See §6.

2. Layered structure (6 physical directories)¶

2.1 Layer table¶

layer	dir	responsibility	may depend on
L1 Interface	`mlmm/cli/`	Click root group, decorator factories, `--help-advanced`, bool flag normalization, subcommand resolver, AmberTools preflight	`workflows/`, `core/`
L2 Application	`mlmm/workflows/`	per-subcommand orchestration plus shared workflow helpers (`_all_helpers.py`, `_opt_freq_common.py`, `_run_session.py`, …)	`domain/`, `backends/`, `io/`, `core/`
L3 Domain	`mlmm/domain/`	chemistry-aware helper logic (bond change detection, bond summary, element-info propagation)	`core/`
L4a Infra (MLIP + ONIOM)	`mlmm/backends/`	MLIP backend dispatch, inline backend integrations, and the ML/MM ONIOM calculator core	`core/`
L4b Infra (I/O)	`mlmm/io/`	output layout, summary, trajectory, PDB fix, energy diagram, Hessian cache, analytical-Hessian glue	`core/`
L5 Foundation	`mlmm/core/`	shared defaults, PDB/XYZ/plot helpers, result commit/output support, and residue tables	`backends/`, `domain/`, `io/` (compatibility back-edges, see below)
(bundle, not a layer)	`<repo>/pysisyphus/`, `<repo>/thermoanalysis/`, `<repo>/hessian_ff/`	repo-internal forks (optimizer / thermochemistry / analytical MM Hessian)	(sibling, layer-external)

Dependency direction: the design intent is one-way L1 → L2 → {L3, L4} → L5. What is actually enforced — and by which checker — is:

.github/scripts/check_import_graph.py (AST import-graph gate) enforces the parts that are true today: (a) the mlmm package has no import cycle (no strongly connected component among its modules); (b) core and domain never import workflows; (c) no bundled fork (pysisyphus / hessian_ff / thermoanalysis) imports mlmm.
.github/scripts/check_engineering_markers.py covers a different concern — chemistry-rule / DOMAIN_PURE marker completeness and the MLIP-runtime import scope. It does not parse the layer import edges, so it does not enforce the direction.

The current cycle-free back-edges are limited to a few core.utils helpers importing domain.add_elem_info and io.structure_formats, plus core.calc_eval importing backends.mlmm_calc. core.utils does not import workflows. Shared charge/spin preparation and layer helpers live in workflows/charge_prep.py and workflows/_opt_freq_common.py. Bundled forks sit outside the layer graph and may be imported from any layer through their absolute package path (from pysisyphus.X import Y, from hessian_ff.analytical_hessian import …).

2.2 ASCII map of the package tree¶

mlmm_toolkit/ [GH: t-0hmura/mlmm_toolkit]
├── pyproject.toml packages.find = ["mlmm*",...] (package-discovery glob)
├── README.md / CONTRIBUTING.md / CHANGELOG.md
├── docs/
│ ├── architecture.md ← this file
│ └──... (Sphinx documentation site)
├── mlmm/ ← package body, 6-layer physical dir
│ ├── __init__.py PEP 562 lazy: _LAZY_IMPORTS + __getattr__
│ ├── __main__.py `from mlmm.cli.app import cli`
│ ├── _version.py / py.typed
│ │
│ ├── cli/ # === L1 Interface ===
│ │ ├── app.py Click group + _LAZY_SUBCOMMANDS registry (absolute paths)
│ │ ├── common_options.py @add_precision_option / @add_backend_model_option / @add_ml_charge_spin_options et al.
│ │ ├── decorators.py make_is_param_explicit, bool/YAML helpers, render_cli_exception
│ │ ├── help_pages.py --help-advanced pager
│ │ ├── bool_compat.py --flag / --no-flag normalization
│ │ ├── default_group.py subcommand resolver, lazy module import
│ │ └── preflight.py AmberTools / conda env / GPU preflight
│ │
│ ├── workflows/ # === L2 Application ===
│ │ ├── all.py full pipeline orchestrator (extract → … → DFT)
│ │ ├── path_search.py / path_opt.py MEP search / COS wrapper
│ │ ├── tsopt.py / freq.py / irc.py / dft.py per-stage runners
│ │ ├── opt.py / scan.py / scan2d.py /
│ │ │ scan3d.py / scan_common.py ONIOM geometry opt / scans
│ │ ├── extract.py active-site extraction CLI
│ │ ├── define_layer.py ML / Movable-MM / Frozen B-factor assignment
│ │ ├── mm_parm.py AmberTools-driven parm7 / rst7 generation
│ │ ├── oniom_export.py ONIOM input writer (Gaussian / ORCA)
│ │ ├── oniom_import.py ONIOM input reader (sanity / atom-name diff)
│ │ ├── align_freeze.py Kabsch + frozen-subset rmsd
│ │ └── _all_helpers.py / _opt_freq_common.py / _run_session.py /
│ │     restraints.py shared workflow helpers
│ │
│ ├── domain/ # === L3 Domain ===
│ │ ├── bond_changes.py R↔P bond detection
│ │ ├── bond_summary.py post-IRC diagnostic
│ │ └── add_elem_info.py PDB element column normalizer
│ │
│ ├── backends/ # === L4a Infra (MLIP + ONIOM) ===
│ │ ├── __init__.py --precision routing (apply_precision_to_calc_cfg)
│ │ ├── mlmm_calc.py ML/MM ONIOM calculator core (4 MLIP backends UMA / ORB / MACE / AIMNet2
│ │ inline; CHEMISTRY-RULE:1 / 2 / 8 / 9 host)
│ │ ├── custom.py user ASE calculator loaded from --calc-file (custom backend)
│ │ └── _determinism.py strict-determinism setup (--deterministic)
│ │
│ ├── io/ # === L4b Infra (I/O) ===
│ │ ├── summary.py summary.json / summary.log writer
│ │ ├── energy_diagram.py Plotly diagram
│ │ ├── trj2fig.py trajectory → PNG / HTML / SVG / PDF
│ │ ├── pdb_fix.py altloc resolution
│ │ ├── hessian_cache.py in-memory Hessian cache
│ │ └── hessian_calc.py numerical-Hessian build + frequency / vibrational I/O helpers
│ │
│ ├── core/ # === L5 Foundation ===
│ │ ├── defaults.py shared workflow/calculator defaults
│ │ ├── utils.py PDB / XYZ / plot helpers
│ │ ├── logging.py -v/--verbose LEVEL (0–3) logging wiring
│ │ ├── calc_eval.py per-stage calc evaluation
│ │ ├── output.py / result_commit.py output/result commit helpers
│ │ ├── pes_composition.py energy-component composition
│ │ └── residue_data.py residue tables
│ │
│ └── mcp/ # non-layer subpackage: MCP server exposing every CLI subcommand
│   ├── server.py / _runner.py
│   └── _tools.py
│
├── tests/ smoke / unit
├── .github/ workflows/ + scripts/ (CI, release, engineering, and docs checks)
└── (repo-top sibling, layer-external bundled forks)
 pysisyphus/ repo-internal optimizer, TS, IRC, COS, and calculator fork
 thermoanalysis/ repo-internal fork
 hessian_ff/ repo-internal native Hessian/MM support, NO upstream PyPI, mandatory bundling

2.3 Per-layer responsibility detail¶

L1 cli/. This layer owns root dispatch and shared argv parsing; leaf Click commands are defined in the registered workflows/, domain/, and io/ modules. app.py holds the root Click.Group plus the _LAZY_SUBCOMMANDS registry — every entry uses an absolute module path (mlmm.workflows.all, mlmm.io.trj2fig, …) so the resolver is independent of where default_group.py itself lives. The mlmm-specific preflight.py (AmberTools / conda env / GPU preflight) lives here because it runs during CLI startup before any L2 workflow is invoked.

L2 workflows/ contains command modules plus shared workflow helpers. Modules registered in cli/app.py:_LAZY_SUBCOMMANDS own a @click.command() named cli; helper modules such as _all_helpers.py, _opt_freq_common.py, _run_session.py, scan_common.py, and restraints.py have no independent command. Large stage runners remain single modules in the current layout.

L3 domain/. Chemistry-aware helper logic that may import torch / numpy / pysisyphus.constants (numeric back-ends), but may not import machine-learning interatomic potential (MLIP) runtimes (fairchem, orb_models, mace, aimnet). Two distinct CI gates cover this, both in .github/scripts/check_engineering_markers.py:

The MLIP-runtime deny list (fairchem / orb_models / mace / aimnet) is enforced repo-wide by _check_external_library_scope, which forbids those imports in any module outside backends/.
The separate # DOMAIN_PURE module-docstring marker is a distinct CI gate (_check_domain_pure) that flags the specific backend-agnostic modules required to stay MLIP-free — backends/mlmm_calc.py, workflows/tsopt.py, workflows/freq.py (and present on workflows/sp.py). It is not itself the deny-list mechanism, and no domain/ file carries it.

Domain helpers are reusable by any L2 stage runner.

L4a backends/. The ML/MM ONIOM calculator core (mlmm_calc.py) lives here together with backend dispatch (__init__.py). The ML-region backends (UMA / ORB / MACE / AIMNet2) and OpenMM / hessian_ff coupling are dispatched from this layer. mlmm_calc.py carries chemistry rules #1 (subtractive ONIOM), #2 (link-atom Hessian B-matrix), and #8 (3-layer 5-pass partial Hessian); rule #9 (parm7 atom indexing) lives in io/pdb_indexing.py — see §5.1.

L4b io/. Output-side I/O concerns include the per-stage summary writer, energy diagram, trajectory rendering, PDB/altloc handling, Hessian cache, numerical Hessian construction, and frequency/vibrational I/O (hessian_calc.py). io/ never depends on workflows/; output format is owned here and consumed by stage runners.

L5 core/. The lowest layer. defaults.py is the single source of truth for shared defaults — grep here before adding a number elsewhere, then inspect justified command-local defaults. utils.py contains shared PDB / XYZ / plotting helpers; logging.py (per-subcommand -v/--verbose LEVEL, 0–3), calc_eval.py (per-stage calculation evaluation), and residue_data.py (residue tables) also live here.

2.4 Lazy-import mechanism (conceptual diagram)¶

External consumer Package root Layer dir
------------------ ---------------- -----------

from mlmm.core.utils import x ────────────────────────────────────► mlmm/core/utils.py

import mlmm.io.trj2fig ──────────────────────────────────────────► mlmm/io/trj2fig.py

from mlmm.backends.mlmm_calc import ─────────────────────────────► mlmm/backends/mlmm_calc.py
 MLMMCore

from mlmm import MLMMCore ─────► mlmm/__init__.py
 __getattr__("MLMMCore")
 └─► _LAZY_IMPORTS["MLMMCore"]
 = "mlmm.backends.mlmm_calc"
 └─► importlib.import_module(...)
 └─► getattr(module, "MLMMCore")

mlmm myaction ─────────────────► mlmm/cli/app.py
 _LAZY_SUBCOMMANDS["myaction"]
 = ("mlmm.workflows.myaction", "cli", "...")
 └─► importlib.import_module(absolute path)
 └─► getattr(module, "cli") → Click command

Two import surfaces are supported:

Layered import path: external code imports directly from the layer directory (see the §2.1 layer table; e.g. from mlmm.backends.mlmm_calc import MLMMCore).
Root symbol attribute (from mlmm import MLMMCore) — handled by mlmm/__init__.py:_LAZY_IMPORTS + PEP 562 __getattr__. The five re-exported symbols (MLMMCore, MLMMASECalculator, mlmm, mlmm_ase, mlmm_mm_only) all resolve to mlmm.backends.mlmm_calc and are loaded on first access, so import mlmm stays cheap (only __version__ is eager). There is no root module-attribute surface — submodules are reached by their full path (import mlmm.io.trj2fig), not as attributes of the top-level package.

The CLI subcommand resolver (cli/app.py:_LAZY_SUBCOMMANDS) uses absolute module paths (e.g. "mlmm.workflows.all") so subcommand discovery is independent of the resolver module’s __package__.

3. Five-step navigation¶

For a contributor opening the repo for the first time, follow this path top-to-bottom; each step closes one concern.

step	minutes	open	what you learn
1	3	`README.md`	one-paragraph elevator pitch + single-command usage
2	5	this file (`docs/architecture.md`) §2 + §4	6-layer dir tree, dependency direction, where each concern lives
3	5	`mlmm/cli/app.py`	Click root group, `_LAZY_SUBCOMMANDS` registry (≈ 22 entries), absolute-path resolution
4	20	`mlmm/workflows/all.py` (skim)	one full subcommand top-to-bottom; trace `extract → mm-parm → ONIOM model → MEP → tsopt → IRC → freq → dft`
5	7	`CONTRIBUTING.md` §3 + §4	5 add-a-X recipes + the “do not touch” hidden constraints

After step 5 you can read any other file by following the file index in §4. The package is intentionally flat within each layer: there is no nested package below mlmm/<layer>/, so product modules are at most two directories below mlmm/.

4. File index — “where does this concern live?”¶

4.1 CLI / entry (L1 `cli/`)¶

concern	file
Click root group + subcommand dispatch	`mlmm/cli/app.py`
Subcommand resolver (lazy import)	`mlmm/cli/default_group.py`
`python -m mlmm` shim	`mlmm/__main__.py`
Shared option decorator factories	`mlmm/cli/common_options.py`
Bool/YAML/exception CLI helpers	`mlmm/cli/decorators.py`
`--help-advanced` pager	`mlmm/cli/help_pages.py`
Bool flag compat (`--flag` / `--no-flag` + value style)	`mlmm/cli/bool_compat.py`
AmberTools / conda env / GPU preflight	`mlmm/cli/preflight.py`

4.2 Workflow stage runners (L2 `workflows/`)¶

Acronyms used below: MEP = minimum-energy path; GSM = growing-string method; COS = chain-of-states; RS-P-RFO = restricted-step partitioned rational-function optimization; RS-I-RFO = restricted-step image-function rational-function optimization; Bofill = the Bofill Hessian-update formula; PHVA = partial Hessian vibrational analysis; IRC = intrinsic reaction coordinate; Kabsch = the Kabsch rigid-body alignment algorithm.

concern	file
Full pipeline orchestrator	`mlmm/workflows/all.py`
Geometry optimization (ONIOM macro/micro pre-opt)	`mlmm/workflows/opt.py`
Scan and 2D/3D energy-landscape grids + shared	`mlmm/workflows/scan{,2d,3d,_common}.py`
MEP search (GSM)	`mlmm/workflows/path_search.py`
MEP optimizer core (pysisyphus COS)	`mlmm/workflows/path_opt.py`
TS optimization (RS-P-RFO / RS-I-RFO / TRIM + Bofill + macro/micro)	`mlmm/workflows/tsopt.py`
Vibrational analysis (PHVA + MLIP active block)	`mlmm/workflows/freq.py`
IRC integration (macro / micro)	`mlmm/workflows/irc.py`
Single-point DFT (gpu4pyscf subprocess, ONIOM-embedded)	`mlmm/workflows/dft.py`
Active-site extraction (cluster cut-out + link-atom cap)	`mlmm/workflows/extract.py`
ML / Movable-MM / Frozen region assignment	`mlmm/workflows/define_layer.py`
AmberTools-driven MM parameter generation	`mlmm/workflows/mm_parm.py`
ONIOM input writer (Gaussian / ORCA)	`mlmm/workflows/oniom_export.py`
ONIOM input reader (sanity, atom-name diff)	`mlmm/workflows/oniom_import.py`
Kabsch / frozen-subset alignment	`mlmm/workflows/align_freeze.py`

4.3 Chemistry helpers (L3 `domain/`)¶

concern	file
R↔P bond change detection	`mlmm/domain/bond_changes.py`
Post-IRC bond summary	`mlmm/domain/bond_summary.py`
PDB element column normalizer	`mlmm/domain/add_elem_info.py`

4.4 MLIP + ONIOM (L4a `backends/`)¶

concern	file
ML/MM ONIOM calculator core + 4 inline MLIP backends + ONIOM coupling	`mlmm/backends/mlmm_calc.py`
`--precision` routing (`apply_precision_to_calc_cfg` / `_PRECISION_DISPATCH`)	`mlmm/backends/__init__.py`
Backend dispatch / factory (`_create_ml_backend`)	`mlmm/backends/mlmm_calc.py`

See MLIP Backends for installation and runtime behavior. Backend implementation changes currently touch mlmm_calc.py and the dispatcher.

4.5 I/O (L4b `io/`)¶

concern	file
`summary.json` / `summary.log` writer	`mlmm/io/summary.py`
Plotly energy diagram	`mlmm/io/energy_diagram.py`
Trajectory → PNG / HTML / SVG / PDF	`mlmm/io/trj2fig.py`
PDB altloc resolution	`mlmm/io/pdb_fix.py`
In-memory Hessian cache (per-run TTL)	`mlmm/io/hessian_cache.py`
Numerical Hessian build + frequency / vibrational I/O	`mlmm/io/hessian_calc.py`
Harmonic restraint setup	`mlmm/workflows/restraints.py` (L2 stage helper)

4.6 Foundation (L5 `core/`)¶

concern	file
Shared workflow and calculator defaults	`mlmm/core/defaults.py`
PDB / XYZ / plot helpers	`mlmm/core/utils.py`
`-v/--verbose LEVEL` (0–3) logging wiring	`mlmm/core/logging.py`
Per-stage calc evaluation	`mlmm/core/calc_eval.py`
Output/result commit helpers	`mlmm/core/output.py`, `mlmm/core/result_commit.py`
Energy-component composition	`mlmm/core/pes_composition.py`
Residue tables	`mlmm/core/residue_data.py`

4.7 Repo-internal bundled forks¶

dir	role	divergent files (do NOT replace with upstream)
`pysisyphus/`	optimizer / TS / IRC engine	`irc/IRC.py`, `optimizers/hessian_updates.py`, `tsoptimizers/TSHessianOptimizer.py`, `calculators/*`
`thermoanalysis/`	thermochemistry (ΔG, ZPE, partition functions)	`QCData.py` (branding diff vs upstream)
`hessian_ff/`	analytical Hessian on MM force field — PyPI 404, bundling is mandatory	`analytical_hessian.py` (sole entry consumed by `mlmm/backends/mlmm_calc.py`)

See each dir’s README.md for the touch-restriction boundary.

5. Scientific invariants¶

5.1 Nine chemistry rules (grep recipe)¶

Nine correctness-critical rules are spread across backends/, workflows, and core/defaults.py. Inline # CHEMISTRY-RULE:N markers and # DOMAIN_PURE module-docstring markers identify their implementation sites; .github/scripts/check_engineering_markers.py checks marker completeness.

To find every chemistry rule before editing:

# List all 9 rule sites in the repo (host file + line)
grep -rnE '# CHEMISTRY-RULE:[0-9]+' mlmm/

# List every # DOMAIN_PURE marker (= chemistry-rule host modules)
grep -rn '# DOMAIN_PURE' mlmm/

All 9 rules apply to mlmm:

#	rule	host file
1	Subtractive ONIOM energy formula (`E = mm_real + ml_model − mm_model`)	`mlmm/backends/mlmm_calc.py`
2	Link-atom Hessian B-matrix projection	`mlmm/backends/mlmm_calc.py`
3	Macro / micro alternation for Hessian TS optimizers (RS-P-RFO default)	`mlmm/workflows/tsopt.py`
4	gpu4pyscf `rks_lowmem` triple-guard	`mlmm/workflows/dft.py`
5	def2 family auto-ECP injection	`mlmm/workflows/dft.py`
6	PHVA + MLIP active-block partial Hessian	`mlmm/workflows/freq.py`
7	`bofill_update` advanced-indexing scatter	`mlmm/workflows/tsopt.py`
8	3-layer 5-pass partial Hessian assembly	`mlmm/backends/mlmm_calc.py`
9	parm7 atom indexing (1-based / serial gap handling)	`mlmm/io/pdb_indexing.py`

Changes to these paths require focused regression tests and the relevant scheduled numerical validation (see CONTRIBUTING.md §1.1).

Recommended learning order (4 chemistry clusters):

cluster	rules	shared concern	learn-first file
5-pass Hessian set	#1, #2, #8, #9	subtractive ONIOM + link-atom B-matrix + 3-layer assembly + parm7 indexing	`mlmm/backends/mlmm_calc.py` (host of 3 of the 9 rules: #1/#2/#8; #9 in `mlmm/io/pdb_indexing.py`)
TS optimization set	#3, #7	macro / micro alternation + Bofill scatter	`mlmm/workflows/tsopt.py`
Vibrational set	#6	PHVA + MLIP active-block partial Hessian	`mlmm/workflows/freq.py`
DFT set	#4, #5	gpu4pyscf low-memory + def2 ECP injection	`mlmm/workflows/dft.py`

For mlmm the practical curriculum is the 5-pass Hessian set first (#1, #2, #8 in mlmm_calc.py; #9 in io/pdb_indexing.py), then the TS set (#3, #7), then DFT (#4, #5), then vibrational (#6).

5.2 VRAM-management invariant (do not refactor `del` chains)¶

The IRC / TSopt / Freq stages explicitly del GPU-resident objects (calc, geom, hess) between stages to free CUDA memory; stage boundaries also use gc.collect() and, where CUDA allocations are present, torch.cuda.empty_cache(). Do not refactor these release operations out — long-running ML/MM all jobs on the full protein environment OOM without them.

5.3 Bundled forks: do NOT install upstream alongside¶

The bundled pysisyphus/, thermoanalysis/, and hessian_ff/ packages are forks (and in the case of hessian_ff/, the only available distribution — PyPI returns 404). Reinstalling pip install pysisyphus or pip install thermoanalysis next to this package silently breaks:

pysisyphus/irc/IRC.py — initial-displacement memory hygiene
pysisyphus/optimizers/hessian_updates.py — GPU-resident in-place rank-two Bofill update; opt-in PYSIS_BOFILL_CPU_OFFLOAD=1 fallback
pysisyphus/tsoptimizers/TSHessianOptimizer.py — Hessian TS optimizer kwargs
pysisyphus/calculators/... — GPU-aware backend hooks
thermoanalysis/QCData.py — branding / I/O diff vs upstream
hessian_ff/analytical_hessian.py — sole entry consumed by backends/mlmm_calc.py; no upstream alternative exists

5.4 Package discovery and runtime dependencies¶

[tool.setuptools.packages.find].include uses the mlmm* glob to discover layer subpackages. A new top-level package layout must therefore be checked against wheel contents, and every imported runtime package must be declared in dependencies.

5.5 `_LAZY_SUBCOMMANDS` registry must use absolute paths¶

mlmm/cli/app.py:_LAZY_SUBCOMMANDS resolves every subcommand through an absolute module path. A relative dotted import (".all" etc.) would make resolution depend on the resolver module’s __package__ instead of the package root.

6. Bundled forks (repo-internal)¶

mlmm_toolkit ships three repo-internal modules at the repo top:

dir	upstream PyPI?	purpose	scope of edits allowed
`pysisyphus/`	NO — fork, do not `pip install pysisyphus` alongside	optimizer, TS, IRC, COS, calculators	preserve the listed divergences; validate numerical changes with focused and scheduled numerical tests
`thermoanalysis/`	NO — fork (branding diff)	ΔG, ZPE, partition functions, `QCData`	preserve the `QCData` consumer contract; see its README
`hessian_ff/`	NO — PyPI 404, bundling mandatory	analytical Hessian on MM force field	preserve its derivative and public API contracts; see its README

Each dir carries its own README.md listing the divergent files and the touch-restriction boundary. From the layer model these forks live outside the L1..L5 graph: any layer may import them via the absolute package path (from pysisyphus.X import Y, from hessian_ff.analytical_hessian import …) without breaking the L1 → L2 → {L3, L4} → L5 direction.

7. Recommended deeper reading order (5–10 files)¶

After the Fresh-eyes tour (§3), follow this depth-first reading order:

mlmm/core/defaults.py — internalise the default-value table; everything downstream reads from here.
mlmm/cli/app.py — Click root + _LAZY_SUBCOMMANDS registry.
mlmm/workflows/all.py — one full pipeline top-to-bottom.
mlmm/workflows/extract.py + define_layer.py — cluster cut-out + link-atom cap + ONIOM layer assignment.
mlmm/workflows/mm_parm.py — AmberTools parm7 generation.
mlmm/backends/mlmm_calc.py — the heart of ML/MM (chemistry-rules #1, #2, #8 live here; #9 in mlmm/io/pdb_indexing.py).
mlmm/workflows/tsopt.py — Hessian TS optimizers + Bofill (CHEMISTRY-RULE:7) + macro / micro alternation (CHEMISTRY-RULE:3).
mlmm/workflows/freq.py — PHVA + MLIP active-block (CHEMISTRY-RULE:6).
mlmm/workflows/irc.py — VRAM hygiene + macro / micro IRC.
mlmm/core/utils.py — shared PDB / XYZ / plot helpers.

8. ML/MM (ONIOM) scope¶

mlmm-toolkit operates on the full protein environment via ONIOM:

ML region: substrate + reaction-center residues, evaluated by one of 4 machine-learning interatomic potential (MLIP) backends (UMA / ORB / MACE / AIMNet2)
Movable-MM region: a shell around the ML region, free to move under the AMBER force field
Frozen region: the rest of the protein, held rigid

The split is encoded in B-factor channels of the input PDB and propagated through extract → mm-parm → ONIOM model → MEP → tsopt → IRC → freq → dft.