mlmm-toolkit 0.3.4__tar.gz → 0.3.5__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/CHANGELOG.md +180 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/CITATION.cff +1 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/PKG-INFO +16 -15
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/README.md +5 -5
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/all.md +20 -11
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/backends.md +3 -3
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/cli-conventions.md +4 -11
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/concepts.md +6 -5
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/device-hpc.md +6 -8
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/dft.md +2 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/extract.md +4 -2
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/freq.md +8 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/getting-started.md +9 -14
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/glossary.md +2 -2
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/irc.md +4 -4
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/all.md +18 -9
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/backends.md +4 -6
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/cli-conventions.md +7 -16
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/concepts.md +7 -6
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/device-hpc.md +7 -11
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/dft.md +2 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/extract.md +4 -3
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/freq.md +7 -0
- mlmm_toolkit-0.3.5/docs/ja/getting-started.md +292 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/glossary.md +3 -3
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/index.md +11 -48
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/irc.md +4 -4
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/json-output.md +22 -9
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/mlmm-calc.md +2 -2
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/opt.md +3 -2
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/output-layout.md +2 -4
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/path-opt.md +1 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/path-search.md +7 -5
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/python-api.md +2 -2
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/quickstart-all.md +3 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/quickstart-scan-spec.md +4 -8
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/quickstart-tsopt-freq.md +2 -2
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/recipes-common-errors.md +3 -3
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/scan.md +12 -12
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/troubleshooting.md +21 -11
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/tsopt.md +45 -15
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/yaml-reference.md +58 -54
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/json-output.md +20 -5
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/mlmm-calc.md +2 -2
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/opt.md +4 -3
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/output-layout.md +1 -3
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/path-opt.md +1 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/path-search.md +7 -5
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/python-api.md +1 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/quickstart-all.md +2 -2
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/quickstart-scan-spec.md +3 -5
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/quickstart-tsopt-freq.md +1 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/recipes-common-errors.md +6 -6
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/all.md +20 -11
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/opt.md +3 -5
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/path_opt.md +1 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/path_search.md +16 -5
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/scan.md +3 -3
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/tsopt.md +3 -6
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/scan.md +11 -9
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/troubleshooting.md +23 -14
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/tsopt.md +42 -12
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/yaml-reference.md +37 -35
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/mlmm_colab.ipynb +293 -76
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/native/loader.py +6 -5
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/_version.py +2 -2
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/backends/mlmm_calc.py +115 -64
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/core/defaults.py +1 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/core/utils.py +6 -3
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/summary.py +122 -17
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/trj2fig.py +1 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/_all_helpers.py +7 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/_outcomes.py +92 -24
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/all.py +471 -114
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/freq.py +14 -7
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/irc.py +23 -6
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/opt.py +16 -3
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/path_opt.py +25 -3
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/path_search.py +180 -19
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/scan.py +4 -2
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/tsopt.py +212 -51
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm_toolkit.egg-info/PKG-INFO +16 -15
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm_toolkit.egg-info/SOURCES.txt +32 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm_toolkit.egg-info/requires.txt +14 -9
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pyproject.toml +13 -9
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/Geometry.py +1 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/normal_modes.py +78 -9
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/HessianOptimizer.py +367 -202
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/RFOptimizer.py +108 -5
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/guess_hessians.py +3 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/hessian_updates.py +72 -31
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/tr_projection.py +6 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/tsoptimizers/RSIRFOptimizer.py +7 -11
- mlmm_toolkit-0.3.5/pysisyphus/tsoptimizers/RSPRFOptimizer.py +505 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/tsoptimizers/TRIM.py +11 -41
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/tsoptimizers/TSHessianOptimizer.py +86 -72
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/SKILL.md +12 -19
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/all-endpoint-mep.md +9 -10
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/all-scan-list.md +3 -4
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/all-ts-only.md +7 -6
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/all.md +10 -10
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/dft.md +7 -4
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/irc.md +16 -2
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/path-search.md +6 -4
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/tsopt.md +1 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-hpc/SKILL.md +6 -8
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-hpc/dynamic-dispatch.md +3 -4
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-install-backends/SKILL.md +12 -11
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-install-backends/core.md +2 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-install-backends/env-cuda.md +6 -7
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-install-backends/mace.md +1 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-install-backends/orb.md +2 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-install-backends/uma.md +4 -5
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-overview/SKILL.md +9 -9
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-structure-io/SKILL.md +1 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-structure-io/xyz.md +3 -3
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-ts-strategy/SKILL.md +3 -3
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-workflows-output/SKILL.md +57 -42
- mlmm_toolkit-0.3.5/tests/test_active_projection_roundtrip.py +115 -0
- mlmm_toolkit-0.3.5/tests/test_aimnet_backend_adapter.py +68 -0
- mlmm_toolkit-0.3.5/tests/test_all_endpoint_failure.py +299 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_all_helpers.py +49 -5
- mlmm_toolkit-0.3.5/tests/test_all_mep_pdb.py +149 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_all_postprocessing_contract.py +10 -2
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_all_summary_provenance.py +1 -1
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_backend_model_load_announcement.py +42 -0
- mlmm_toolkit-0.3.5/tests/test_bfgs_indefinite_routing.py +151 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_bundled_pysis_policy.py +1 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_bundled_pysisyphus_regressions.py +7 -5
- mlmm_toolkit-0.3.5/tests/test_cartesian_atomic_trust.py +339 -0
- mlmm_toolkit-0.3.5/tests/test_cartesian_ts_defaults.py +255 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_cli_feature_regressions.py +65 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_cli_help.py +2 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_colab_notebook.py +398 -24
- mlmm_toolkit-0.3.5/tests/test_complete_physical_modes.py +103 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_cycle_limit_defaults.py +1 -1
- mlmm_toolkit-0.3.5/tests/test_dependency_metadata.py +62 -0
- mlmm_toolkit-0.3.5/tests/test_embed_micro_dispatch.py +260 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_energy_plateau_stall.py +9 -4
- mlmm_toolkit-0.3.5/tests/test_fairchem_device_ownership.py +399 -0
- mlmm_toolkit-0.3.5/tests/test_frequency_cli_thermochemistry.py +117 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_frequency_zero_cutoff.py +30 -2
- mlmm_toolkit-0.3.5/tests/test_geometry_rebuild_cache.py +128 -0
- mlmm_toolkit-0.3.5/tests/test_hessian_update_boundaries.py +302 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_mep_threshold_flags.py +33 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_microiteration_cycle_carry.py +76 -3
- mlmm_toolkit-0.3.5/tests/test_optimizer_curvature.py +327 -0
- mlmm_toolkit-0.3.5/tests/test_optimizer_partial_curvature.py +141 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_optimizer_trial_safeguards.py +19 -7
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_path_opt_output_boundary.py +13 -0
- mlmm_toolkit-0.3.5/tests/test_path_search_max_depth.py +113 -0
- mlmm_toolkit-0.3.5/tests/test_phva_physical_oracle.py +119 -0
- mlmm_toolkit-0.3.5/tests/test_rfo_micro_boundary.py +114 -0
- mlmm_toolkit-0.3.5/tests/test_rfo_model_descent.py +191 -0
- mlmm_toolkit-0.3.5/tests/test_rfo_symmetric.py +277 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_scientific_outcomes.py +178 -54
- mlmm_toolkit-0.3.5/tests/test_secular_near_pole.py +102 -0
- mlmm_toolkit-0.3.5/tests/test_secular_scale_contract.py +86 -0
- mlmm_toolkit-0.3.5/tests/test_secular_working_precision.py +45 -0
- mlmm_toolkit-0.3.5/tests/test_small_radius_growth.py +69 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_smoke_cli_contract.py +1 -1
- mlmm_toolkit-0.3.5/tests/test_soft_curvature_trajectory.py +221 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_summary_log.py +95 -2
- mlmm_toolkit-0.3.5/tests/test_trim_image_step.py +211 -0
- mlmm_toolkit-0.3.5/tests/test_trust_quality_validity.py +238 -0
- mlmm_toolkit-0.3.5/tests/test_trust_region_step.py +49 -0
- mlmm_toolkit-0.3.5/tests/test_ts_terminal_cadence.py +170 -0
- mlmm_toolkit-0.3.5/tests/test_ts_terminal_screen.py +239 -0
- mlmm_toolkit-0.3.5/tests/test_ts_translation_filter.py +69 -0
- mlmm_toolkit-0.3.5/tests/test_tsopt_terminal_phva.py +355 -0
- mlmm_toolkit-0.3.5/tests/test_tsopt_terminal_workflow_phva.py +180 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_write_result_json.py +2 -1
- mlmm_toolkit-0.3.4/docs/ja/getting-started.md +0 -558
- mlmm_toolkit-0.3.4/pysisyphus/tsoptimizers/RSPRFOptimizer.py +0 -211
- mlmm_toolkit-0.3.4/tests/test_dependency_metadata.py +0 -23
- mlmm_toolkit-0.3.4/tests/test_tsopt_terminal_phva.py +0 -168
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/LICENSE +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/MANIFEST.in +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/THIRD_PARTY_NOTICES.txt +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/add-elem-info.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/architecture.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/bond-summary.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/cif.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/define-layer.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/energy-diagram.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/fix-altloc.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/index.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/add-elem-info.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/architecture.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/bond-summary.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/cif.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/define-layer.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/energy-diagram.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/fix-altloc.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/mcp_server.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/mm-parm.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/oniom-export.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/oniom-gaussian.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/oniom-import.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/oniom-orca.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/reproducibility.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/scan2d.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/scan3d.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/sp.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/ja/trj2fig.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/mcp_server.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/mm-parm.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/oniom-export.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/oniom-gaussian.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/oniom-import.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/oniom-orca.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/add_elem_info.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/bond_summary.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/define_layer.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/dft.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/energy_diagram.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/extract.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/fix_altloc.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/freq.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/index.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/irc.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/mm_parm.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/oniom_export.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/oniom_import.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/scan2d.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/scan3d.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/sp.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/commands/trj2fig.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reference/yaml.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/reproducibility.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/scan2d.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/scan3d.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/sp.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/docs/trj2fig.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/beza/1.R.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/beza/2.IM.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/beza/3.P.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/beza/README.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/beza/run.sh +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/comt/1.R.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/comt/3.P.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/comt/README.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/comt/run.sh +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/mcp_client_config.json +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/methyltransferase/complex.parm7 +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/methyltransferase/complex.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/methyltransferase/complex.rst7 +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/methyltransferase/pocket.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/methyltransferase/r_layered.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/methyltransferase/run_all.sh +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/methyltransferase/run_stepwise.sh +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/toy_system/ml_region_p.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/toy_system/ml_region_r.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/toy_system/opt_ase.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/toy_system/p_complex.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/toy_system/p_complex.rst7 +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/toy_system/p_toy.parm7 +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/toy_system/p_toy.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/toy_system/r_complex.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/toy_system/r_toy.pdb +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/toy_system/run.sh +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/examples/toy_system/test_core.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/active_space.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/analytical_hessian.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/constants.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/forcefield.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/loaders.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/native/Makefile +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/native/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/native/analytical_hessian.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/native/analytical_hessian_ext.cpp +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/native/bonded.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/native/bonded_ext.cpp +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/native/nonbonded.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/native/nonbonded_ext.cpp +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/prmtop_parmed.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/system.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/terms/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/terms/angle.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/terms/bond.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/terms/cmap.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/terms/dihedral.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/terms/nonbonded.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/tests/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/tests/conftest.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/tests/test_active_space.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/tests/test_coords_input.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/tests/test_energy_force.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/tests/test_hessian.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/tests/test_input_boundaries.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/tests/test_native_loader.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/tests/test_runtime_cache.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/tests/test_smoke.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/tests/test_validation.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/hessian_ff/workflows.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/__main__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/backends/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/backends/_determinism.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/backends/custom.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/backends/methods.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/backends/xtb_embedcharge_correction.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/cli/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/cli/app.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/cli/bool_compat.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/cli/common_options.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/cli/decorators.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/cli/default_group.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/cli/help_pages.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/cli/preflight.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/core/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/core/calc_eval.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/core/logging.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/core/output.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/core/pes_composition.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/core/residue_data.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/core/result_commit.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/domain/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/domain/add_elem_info.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/domain/bond_changes.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/domain/bond_summary.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/altloc.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/energy_diagram.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/gaussian_input.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/hessian_cache.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/hessian_calc.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/hessian_file.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/oniom_identity.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/path_mode_cache.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/pdb_fix.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/pdb_indexing.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/plotly_image.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/structure_formats.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/io/xyz_trajectory.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/mcp/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/mcp/_runner.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/mcp/_tools.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/mcp/server.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/py.typed +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/_microiteration.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/_opt_freq_common.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/_run_session.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/align_freeze.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/charge_prep.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/define_layer.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/dft.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/extract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/mm_parm.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/oniom_export.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/oniom_import.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/restraints.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/scan2d.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/scan3d.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/scan_common.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm/workflows/sp.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm_toolkit.egg-info/dependency_links.txt +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm_toolkit.egg-info/entry_points.txt +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/mlmm_toolkit.egg-info/top_level.txt +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/LICENSE +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/TablePrinter.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/_array.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/calculators/Calculator.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/calculators/Dimer.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/calculators/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/config.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/constants.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/cos/ChainOfStates.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/cos/GrowingChainOfStates.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/cos/GrowingString.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/cos/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/elem_data.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/helpers.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/helpers_pure.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/Bend.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/Bend2.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/BondedFragment.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/Cartesian.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/CartesianCoords.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/Coords.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/DLC.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/DistanceFunction.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/DummyImproper.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/DummyTorsion.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/LinearBend.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/LinearDisplacement.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/OutOfPlane.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/PrimTypes.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/Primitive.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/RedundantCoords.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/RobustTorsion.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/Rotation.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/Stretch.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/Torsion.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/Torsion2.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/Translation.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/augment_bonds.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/derivatives.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/eval.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/exceptions.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/helpers.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/logging_conf.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/setup.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/setup_fast.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/update.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/intcoords/valid.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/io/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/io/cjson.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/io/crd.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/io/hdf5.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/io/hessian.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/io/mol2.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/io/pdb.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/io/qcschema.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/io/xyz.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/io/zmat.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/irc/DWI.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/irc/EulerPC.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/irc/IRC.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/irc/Instanton.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/irc/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/linalg.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/modefollow/NormalMode.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/modefollow/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/modefollow/lanczos.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/BacktrackingOptimizer.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/LBFGS.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/Optimizer.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/StringOptimizer.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/checkpoint.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/closures.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/exceptions.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/gdiis.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/poly_fit.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/optimizers/restrict_step.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/stocastic/FragmentKick.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/stocastic/Kick.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/stocastic/Pipeline.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/stocastic/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/stocastic/align.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/tsoptimizers/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/pysisyphus/xyzloader.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/setup.cfg +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/README.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/colab-local-gpu-runtime/SKILL.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/colab-local-gpu-runtime/agents/openai.yaml +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/colab-local-gpu-runtime/references/colab-connection.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/colab-local-gpu-runtime/scripts/setup.ps1 +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/colab-local-gpu-runtime/scripts/start.ps1 +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/colab-local-gpu-runtime/scripts/status.ps1 +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/colab-local-gpu-runtime/scripts/stop.ps1 +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-architecture/SKILL.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/add-elem-info.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/bond-summary.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/define-layer.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/energy-diagram.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/extract.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/fix-altloc.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/freq.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/mm-parm.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/oniom-export.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/oniom-import.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/opt.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/path-opt.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/scan.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/scan2d.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/scan3d.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/sp.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-cli/trj2fig.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-env-detect/SKILL.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-install-backends/aimnet2.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-install-backends/ambertools.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-install-backends/dft.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-install-backends/xtb.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-mcp/SKILL.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-structure-io/charge-multiplicity.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-structure-io/cif.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-structure-io/gjf.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-structure-io/parm7.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/skills/mlmm-structure-io/pdb.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/README.md +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/conftest.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_active_map_invalidation.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_add_elem_info_output.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_advanced_help_dedup.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_align_log.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_all_calculator_template.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_all_dry_run_semantics.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_all_layer_honor.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_all_ml_region_subset.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_all_multiplicity_option.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_all_path_tangent.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_all_run_ownership.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_all_scan_contract_checker.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_array_shim_equivalence.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_beza_example_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_bond_changes.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_bool_compat_cli.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_calc_eval.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_charge_prep_exact_model.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_cli_utils.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_confirmed_runtime_boundaries.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_custom_calc_file.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_default_group.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_defaults.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_determinism.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_dft_model_parm7_lj_tables.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_dft_output_boundary.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_dft_result_payload.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_dimer_constraints.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_dmf_result_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_docs_command_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_docs_command_smoke.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_docs_deploy_workflow.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_docs_link_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_dry_run_footer_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_dynamic_dispatch_skill.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_e2e_minimal.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_embedcharge_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_error_envelope.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_error_handling.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_example_scripts_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_extract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_freeze_noncart_geometry.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_freq_active_dof.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_gaussian_input.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_generated_reference_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_gpu_kernel_parity.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_heavy_ts_cycle_budget.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_help_grouping.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_hessian_basis_reconciliation.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_hessian_cache_handoff.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_hessian_calc_boundaries.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_hessian_file.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_hessian_method_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_hessian_optimizer_internal_partial.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_hessian_writeback.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_import_graph_gate.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_imports.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_inline_xyz.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_irc_active_space.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_irc_endpoint_hessian_guard.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_irc_never_stop.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_layer_index_validation.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_mace_anicc_factory.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_mcp_runner.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_mcp_tools.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_micro_force_equilibrium.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_microiteration_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_ml_layer_precedence.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_ml_region_outputs.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_mlmm_calc_serial_gap.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_mlmm_result_constraints.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_mm_parm_preflight.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_normal_modes_lowering.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_oniom_export_regressions.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_opt.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_opt_flatten_mass_unweight.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_optimizer_checkpoint.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_orb_analytical_hessian_guard.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_output_adapter.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_path_logging_sections.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_path_semantics.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_pdb_fix.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_pdb_indexing_resolver.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_plotly_image_timeout.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_prealign_activation_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_precision_dispatch.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_preflight.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_reject_uphill_flag.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_release_metadata_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_release_workflow.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_residue_catalog_parity.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_restrained_pes_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_result_commit.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_rfo_uphill_rejection.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_run_session.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_scan3d_csv_provenance.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_scan_energy_accounting.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_scan_oniom_smoke.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_scan_precedence.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_scan_validation_regressions.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_silent_noop_warnings.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_skill_command_checker.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_skill_drift_checker.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_sp_ml_region_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_stop_plateau_flag.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_structure_formats.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_symmetrize.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_thermo_policy.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_thermo_symmetry_number.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_thermochemistry_golden.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_toy_system_contract.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_tr_projection.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_tr_projection_default.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_tsopt_microiter_modes.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_tsopt_output_boundary.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_tsopt_probe_core_lifetime.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_tsopt_reference_mode.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_tsopt_soft_imaginary_warning.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_utils.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_workers_option.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_xtb_embedcharge_parsers.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/tests/test_xyz_trajectory.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/thermoanalysis/LICENSE +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/thermoanalysis/QCData.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/thermoanalysis/__init__.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/thermoanalysis/config.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/thermoanalysis/constants.py +0 -0
- {mlmm_toolkit-0.3.4 → mlmm_toolkit-0.3.5}/thermoanalysis/thermo.py +0 -0
|
@@ -8,6 +8,186 @@ The format follows [Keep a Changelog](https://keepachangelog.com/).
|
|
|
8
8
|
|
|
9
9
|
_No changes yet._
|
|
10
10
|
|
|
11
|
+
## [0.3.5] — 2026-09-16
|
|
12
|
+
|
|
13
|
+
> Upgrade warning: unchanged inputs can produce a different `scientific_status` for
|
|
14
|
+
> standalone `irc` and for `all` (including `failed` where a run with no usable
|
|
15
|
+
> required output previously read `partial`), a different `tsopt` `status` when
|
|
16
|
+
> the final TS energy is unavailable, a different DMF `status` / `converged` for
|
|
17
|
+
> IPOPT code 1, a looser DMF iteration cap, and one subdivision level fewer for a
|
|
18
|
+
> configured `search.max_depth`; `result.json` / `summary.json` now carry
|
|
19
|
+
> `schema_version: "3.0"`. Users of those files must review the Breaking changes
|
|
20
|
+
> below before upgrading.
|
|
21
|
+
|
|
22
|
+
> Cartesian optimizer trajectories and convergence verdicts can also change.
|
|
23
|
+
> Frequency exports now retain the complete signed physical spectrum;
|
|
24
|
+
> thermochemical corrections can change when soft positive modes that were
|
|
25
|
+
> previously omitted are restored. Resolved imaginary-mode reporting remains
|
|
26
|
+
> separate from the raw spectrum and optimizer curvature checks.
|
|
27
|
+
|
|
28
|
+
### Breaking changes
|
|
29
|
+
- **JSON schema 3.0 (breaking).** Remove the IRC `forward_converged` /
|
|
30
|
+
`backward_converged` keys from `irc` `result.json` and from the `all`
|
|
31
|
+
`post_segments[].irc` relay. They reported the conjunction of
|
|
32
|
+
`*_integration_converged` and `*_downhill_departure_valid` under a name that
|
|
33
|
+
read as an IRC success verdict. Both components remain published, and
|
|
34
|
+
`*_integration_converged` is now documented. Gate usability on `*_status`.
|
|
35
|
+
- **IRC directional status semantics (breaking).** A requested direction with a
|
|
36
|
+
finite downhill trajectory and no numerical propagation failure is now usable,
|
|
37
|
+
so an ordinary energy-rise stop, a max-cycle stop, and `--never-stop` report
|
|
38
|
+
`*_status: "stopped"` and can yield `scientific_status: "success"` where the
|
|
39
|
+
same input previously produced `partial`.
|
|
40
|
+
- **`all` endpoint acceptance (breaking).** Final reactant/product acceptance now
|
|
41
|
+
requires both endpoint-optimization convergence flags plus bond-topology
|
|
42
|
+
validation of the optimized endpoints against their assigned MEP endpoints. A
|
|
43
|
+
run whose `post_segments[].endpoint_opt` record is absent or whose optimized
|
|
44
|
+
topology does not match no longer reports `success`.
|
|
45
|
+
- **`tsopt` no longer reports `converged` without a final energy (breaking).** A
|
|
46
|
+
failed final-energy evaluation recorded `status: "converged"` beside
|
|
47
|
+
`energy_hartree: null`; `status` is now `"energy_missing"`.
|
|
48
|
+
`optimization_status` keeps the optimizer's own verdict, so the two fields are
|
|
49
|
+
no longer always equal, and `all`'s TS-to-IRC decision reads it unchanged.
|
|
50
|
+
- **An unusable required output is `failed`, not `partial` (breaking).** A
|
|
51
|
+
required leaf's own retained trajectory no longer promotes a run; only a
|
|
52
|
+
diagnostic leaf's artifact can. A run with no usable required output now reads
|
|
53
|
+
`scientific_status: "failed"` with `execution_status: "completed"` — an IRC
|
|
54
|
+
whose every direction hard-failed, an endpoint-HEI path, or a nonconverged MEP.
|
|
55
|
+
- **`all` reports a requested post stage that never ran (breaking).** With no
|
|
56
|
+
reactive segment, `--tsopt` / `--thermo` / `--dft` were skipped under a
|
|
57
|
+
reasonless `success`. It is now `partial`, and `pipeline_stop` records
|
|
58
|
+
`post` / `no_reactive_segment`.
|
|
59
|
+
- **DMF convergence uses one definition (breaking).** `path-opt` mapped the IPOPT
|
|
60
|
+
status with `status == 0` for `status` / `converged` while its stage leaf used
|
|
61
|
+
`ipopt_status_to_converged`, so code 1 (`Solved_To_Acceptable_Level`) read
|
|
62
|
+
`not_converged` beside `reason: "ipopt_converged"`. Both use the shared mapper
|
|
63
|
+
now, and code 1 is `converged`.
|
|
64
|
+
- **`search.max_depth` counts subdivision levels (breaking).** The cap is now
|
|
65
|
+
`depth >= max_depth`, so the value is the number of recursive subdivision
|
|
66
|
+
levels allowed and `0` performs none, returning each input pair as one MEP
|
|
67
|
+
segment (none when its HEI sits at an endpoint). A
|
|
68
|
+
configured `search.max_depth: N` therefore subdivides one level less than
|
|
69
|
+
before; the default 10 permits 10 levels, previously 11.
|
|
70
|
+
|
|
71
|
+
### Added
|
|
72
|
+
- Add IRC `forward_status` / `backward_status` (`stopped`, `failed`, `disabled`),
|
|
73
|
+
`*_requested`, and `*_integration_stop_reason` to `result.json`, and document the
|
|
74
|
+
already-published `*_integration_converged`.
|
|
75
|
+
- Add TSOPT `opt_mode_requested` / `optimizer` and scan `scan_opt_mode` /
|
|
76
|
+
`scan_optimizer`.
|
|
77
|
+
- Add `post_segments[].endpoint_opt.connectivity_validated` with the optimized
|
|
78
|
+
reactant/product bond-topology record.
|
|
79
|
+
- Add the advanced `--max-depth` option to `path-search` and to `all --refine-path`,
|
|
80
|
+
exposing the recursive-subdivision level cap that was previously YAML-only.
|
|
81
|
+
- Add path-search `search_max_depth`, IRC `forward_short_branch` /
|
|
82
|
+
`backward_short_branch`, `all` `config.preopt`, and `Pipeline stop` /
|
|
83
|
+
`Recursion depth cap` lines in `summary.log`.
|
|
84
|
+
|
|
85
|
+
### Changed
|
|
86
|
+
|
|
87
|
+
- For ordinary Cartesian RS-P-RFO, default to TS-BFGS updates and a 0.1 Å
|
|
88
|
+
maximum-atom initial/maximum trust radius when those settings are omitted.
|
|
89
|
+
Explicit optimizer settings remain authoritative.
|
|
90
|
+
- With BFGS selected, use the existing TS-BFGS update for an indefinite
|
|
91
|
+
working Hessian while retaining the nonpositive-curvature skip. Keep a
|
|
92
|
+
descending RFO reference step when an extrapolated step is non-descending.
|
|
93
|
+
- Continue no-flatten RS-P-RFO from higher-order candidates with fresh
|
|
94
|
+
curvature checks; preserve explicitly requested flatten workflows.
|
|
95
|
+
- Preserve scheduled Hessian refreshes during no-flatten RS-P-RFO continuation.
|
|
96
|
+
- Document startup BLIS thread settings for slow DMF/IPOPT calculations.
|
|
97
|
+
- Support fairchem-core 2.22 and current compatible runtime dependencies.
|
|
98
|
+
- Raise the default DMF IPOPT iteration cap (`--max-cycles-dmf`,
|
|
99
|
+
`dmf.max_cycles`) from 300 to 3000, matching the solver's own default. A DMF
|
|
100
|
+
MEP that previously stopped at that cap can now continue.
|
|
101
|
+
- `all` `config.ts_opt_mode` / `config.endpoint_opt_mode` are `null` when TS
|
|
102
|
+
optimization was not requested, instead of always carrying the default preset.
|
|
103
|
+
|
|
104
|
+
### Fixed
|
|
105
|
+
- Repair ORB installation on Python 3.13 Colab runtimes by preparing its
|
|
106
|
+
pinned dm-tree build dependency, while preserving the resolved Torch and NumPy
|
|
107
|
+
versions and exposing installer diagnostics.
|
|
108
|
+
- Normalize quoted scan inputs in the Colab command display.
|
|
109
|
+
- Prefer PDB structures and trajectories in Colab Results, preserving atom,
|
|
110
|
+
residue and chain metadata and all trajectory frames. Link XYZ energies only
|
|
111
|
+
to matching current-run PDB frames; retain XYZ fallback and signed mode labels.
|
|
112
|
+
- Address the UMA first-call CUDA device mismatch reported in
|
|
113
|
+
[pdb2reaction #298](https://github.com/t-0hmura/pdb2reaction/pull/298) by keeping
|
|
114
|
+
input batches on the CPU during FAIR-Chem's lazy initialization, before its
|
|
115
|
+
native device transfer. Add regression coverage for initial and repeated
|
|
116
|
+
energy/force evaluations.
|
|
117
|
+
- Retain all signed physical frequency modes and their vectors, including positive
|
|
118
|
+
modes at or below the reporting cutoff, in frequency exports and thermochemistry.
|
|
119
|
+
Report resolved imaginary counts separately from all negative modes; near-zero
|
|
120
|
+
metadata now annotates the complete spectrum. Thermal corrections can change
|
|
121
|
+
when previously omitted positive modes are restored. Existing thermal policies,
|
|
122
|
+
reporting thresholds and strict curvature acceptance remain unchanged.
|
|
123
|
+
- Avoid allocating an unused square left-singular-vector matrix when many
|
|
124
|
+
frozen atoms constrain the rigid-mode subspace.
|
|
125
|
+
- Keep optimizer mode indices out of flatten-branch scoring after the final
|
|
126
|
+
frequency analysis expands to a different active space.
|
|
127
|
+
|
|
128
|
+
- Include near-zero negative modes in OPT/TS curvature acceptance while keeping
|
|
129
|
+
frequency display thresholds unchanged.
|
|
130
|
+
- Clear cached results when internal-coordinate rebuilding changes the geometry.
|
|
131
|
+
- Handle remaining negative image curvature in TRIM's trust-region step.
|
|
132
|
+
- Honor small trust radii in RFO and RS-I-RFO microiterations.
|
|
133
|
+
- Use standard optimization with embedding, since MM-only microiterations omit embedding forces.
|
|
134
|
+
- Write `all` MEP PDB trajectories when XYZ coordinates have a separate reference PDB.
|
|
135
|
+
- Keep UMA's backbone in evaluation mode during analytical Hessian calculation.
|
|
136
|
+
- Handle zero or non-finite energy predictions consistently when updating trust radii.
|
|
137
|
+
- Solve scaled RFO with an equivalent symmetric matrix and reject unresolved
|
|
138
|
+
restricted steps while preserving terminal saddle checks and recovery.
|
|
139
|
+
- Handle degenerate trust-region steps and keep TS roots in the active Hessian basis.
|
|
140
|
+
- Reject inaccurate secular RFO roots and avoid growing small trust radii after
|
|
141
|
+
interior steps.
|
|
142
|
+
- Require a current-coordinate curvature check before accepting an RFO minimum.
|
|
143
|
+
- Preserve genuine complementary negative curvature after earlier TS validation;
|
|
144
|
+
restrict artifact stabilization to pure, unconstrained translations.
|
|
145
|
+
- Reuse terminal frequencies only for matching analysis conditions, preserving
|
|
146
|
+
partial-Hessian ordering and separating macro-space from final ML/MM validation.
|
|
147
|
+
- Retain microiteration failure diagnostics and honor the CLI flatten setting.
|
|
148
|
+
- Handle singular BFGS/TS-BFGS inputs and reset multistep Hessian history when
|
|
149
|
+
rebuilding the coordinate basis.
|
|
150
|
+
- Stop frequency/DFT and refined diagrams after endpoint execution failures;
|
|
151
|
+
retain diagnostic structures and failure details.
|
|
152
|
+
- Reject fewer than two GSM internal nodes during `path-opt` / `path-search`
|
|
153
|
+
preflight, including dry runs and YAML configuration.
|
|
154
|
+
- Clarify input roles, TS validation, and GSM/DMF help; shorten installation
|
|
155
|
+
guidance and repeated introductory documentation.
|
|
156
|
+
- Record executed single-structure methods in `path_optimizers`, preserving
|
|
157
|
+
citations for scan, alignment, and recursive refinement with `--no-preopt`.
|
|
158
|
+
- Retain a refined path whose HEI is at an endpoint as raw output, without
|
|
159
|
+
reporting it as a reactive segment.
|
|
160
|
+
- Match compiler guidance and job preflight checks to the C++20 flag selected
|
|
161
|
+
by current PyTorch JIT builds.
|
|
162
|
+
- Relay `preopt` consistently into path-search citation payloads: with
|
|
163
|
+
`--no-preopt`, `summary.json`, `summary.log`, and stdout no longer cite an
|
|
164
|
+
optimizer the run never used.
|
|
165
|
+
- Carry `pipeline_stop` into the `summary.log` payload; only the TS-only route
|
|
166
|
+
supplied the key, so the new line never rendered for a path run that stopped
|
|
167
|
+
early.
|
|
168
|
+
- Publish a segment record for an interval that abandoned recursion, and reject
|
|
169
|
+
one whose HEI sits at an endpoint. `n_segments` / `segments[]` had omitted the
|
|
170
|
+
interval and its barrier, ΔE, bond changes and convergence. A failed
|
|
171
|
+
bond-change evaluation now carries a sentinel instead of an empty summary,
|
|
172
|
+
which read as "no covalent change" and dropped the interval from
|
|
173
|
+
post-processing.
|
|
174
|
+
- Fail closed on a missing `tsopt` post-segment record when TSOPT was requested,
|
|
175
|
+
and give `tsopt_missing` / `endpoint_opt_missing` messages.
|
|
176
|
+
- Report `endpoint_opt.connectivity.method` from the probe's own resolution
|
|
177
|
+
instead of always claiming `bond_topology`.
|
|
178
|
+
- Match a suffixed segment's MEP scratch directory when claiming its endpoint
|
|
179
|
+
trajectory, so a depth-capped segment keeps its IRC endpoint orientation.
|
|
180
|
+
- Reject `--max-depth` without `--refine-path`, which was accepted, dropped, and
|
|
181
|
+
still echoed.
|
|
182
|
+
- Report a missing IRC direction record as `failed`, not `disabled`.
|
|
183
|
+
- Restore AIMNet2 evaluation and static image export with current releases.
|
|
184
|
+
- Report TS/IRC-endpoint optimizer modes and requested/effective TS optimizer JSON separately.
|
|
185
|
+
- Show the formatted result warning, including its recovery guidance such as
|
|
186
|
+
`--flatten` for a TS with more than one imaginary mode, in the Colab alert
|
|
187
|
+
instead of the raw status code.
|
|
188
|
+
- Derive the Colab per-direction IRC mark from `*_status` so it can no longer
|
|
189
|
+
contradict the headline scientific status.
|
|
190
|
+
|
|
11
191
|
## [0.3.4] — 2026-09-02
|
|
12
192
|
|
|
13
193
|
### Added
|
|
@@ -12,7 +12,7 @@ authors:
|
|
|
12
12
|
license: GPL-3.0-or-later
|
|
13
13
|
repository-code: "https://github.com/t-0hmura/mlmm_toolkit"
|
|
14
14
|
url: "https://t-0hmura.github.io/mlmm_toolkit/"
|
|
15
|
-
version: "0.3.
|
|
15
|
+
version: "0.3.5"
|
|
16
16
|
preferred-citation:
|
|
17
17
|
type: article
|
|
18
18
|
title: "ML/MM toolkit – Toward Accelerated Mechanistic Investigation of Enzymatic Reactions"
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.4
|
|
2
2
|
Name: mlmm-toolkit
|
|
3
|
-
Version: 0.3.
|
|
3
|
+
Version: 0.3.5
|
|
4
4
|
Summary: mlmm - ML/MM toolkit for enzyme reaction analysis
|
|
5
5
|
Author-email: Takuto Ohmura <ohmura.takuto@gmail.com>
|
|
6
6
|
License-Expression: GPL-3.0-or-later
|
|
@@ -24,14 +24,14 @@ License-File: THIRD_PARTY_NOTICES.txt
|
|
|
24
24
|
License-File: pysisyphus/LICENSE
|
|
25
25
|
License-File: thermoanalysis/LICENSE
|
|
26
26
|
Requires-Dist: fairchem-core
|
|
27
|
-
Requires-Dist: torch
|
|
27
|
+
Requires-Dist: torch
|
|
28
28
|
Requires-Dist: ninja
|
|
29
|
-
Requires-Dist: numpy
|
|
30
|
-
Requires-Dist: biopython
|
|
31
|
-
Requires-Dist: ase
|
|
29
|
+
Requires-Dist: numpy
|
|
30
|
+
Requires-Dist: biopython
|
|
31
|
+
Requires-Dist: ase
|
|
32
32
|
Requires-Dist: ParmEd>=4.3.1
|
|
33
|
-
Requires-Dist: pyyaml
|
|
34
|
-
Requires-Dist: click
|
|
33
|
+
Requires-Dist: pyyaml
|
|
34
|
+
Requires-Dist: click
|
|
35
35
|
Requires-Dist: plotly>=6.1.1
|
|
36
36
|
Requires-Dist: kaleido
|
|
37
37
|
Requires-Dist: autograd
|
|
@@ -47,9 +47,10 @@ Requires-Dist: sympy
|
|
|
47
47
|
Requires-Dist: scikit-learn
|
|
48
48
|
Requires-Dist: pyparsing
|
|
49
49
|
Requires-Dist: pandas
|
|
50
|
-
Requires-Dist: pymsym
|
|
50
|
+
Requires-Dist: pymsym
|
|
51
51
|
Provides-Extra: orb
|
|
52
|
-
Requires-Dist: orb-models; extra == "orb"
|
|
52
|
+
Requires-Dist: orb-models>=0.7.0; python_version >= "3.12" and extra == "orb"
|
|
53
|
+
Requires-Dist: orb-models<0.6,>=0.5.5; python_version < "3.12" and extra == "orb"
|
|
53
54
|
Provides-Extra: aimnet
|
|
54
55
|
Requires-Dist: aimnet>=0.2.0; extra == "aimnet"
|
|
55
56
|
Provides-Extra: pdbfixer
|
|
@@ -72,7 +73,7 @@ Requires-Dist: myst-parser>=2.0; extra == "docs"
|
|
|
72
73
|
Requires-Dist: furo>=2024.0; extra == "docs"
|
|
73
74
|
Requires-Dist: sphinx-copybutton>=0.5; extra == "docs"
|
|
74
75
|
Provides-Extra: mcp
|
|
75
|
-
Requires-Dist: mcp[cli]
|
|
76
|
+
Requires-Dist: mcp[cli]<2,>=1.29; extra == "mcp"
|
|
76
77
|
Dynamic: license-file
|
|
77
78
|
|
|
78
79
|
# **mlmm-toolkit**: An End-to-End ML/MM ONIOM Platform for Automated Enzymatic Reaction Mechanism Analysis
|
|
@@ -125,18 +126,18 @@ The source repository includes full-system COMT and [BezA](examples/beza/README.
|
|
|
125
126
|
| GPU / CUDA / VRAM | A backend-compatible NVIDIA GPU/driver for GPU execution; size VRAM from a representative target-system pilot. |
|
|
126
127
|
| RAM / Disk | Size RAM and disk for the selected backend, model cache, topology tools, and expected artifacts. |
|
|
127
128
|
|
|
128
|
-
|
|
129
|
+
**AmberTools** (`tleap`) is required to generate a topology with `mm-parm` or `all`; reuse an existing matching `--parm` to skip that preparation. The default `hessian_ff` backend needs **a C++20-capable compiler** (validated with GCC 13.3) to JIT-compile native kernels on first use. **pdbfixer** is optional — only `mm-parm --add-h` needs it — `conda install -c conda-forge ambertools pdbfixer -y` installs both. CPU-only ML/MM execution is supported but can be substantially slower than GPU execution; benchmark the selected backend and system. Full requirement and tuning details: [docs/getting-started.md#installation](docs/getting-started.md#installation).
|
|
129
130
|
|
|
130
131
|
## Installation
|
|
131
132
|
|
|
132
133
|
```bash
|
|
133
134
|
# 1. New env + AmberTools + CUDA-enabled PyTorch
|
|
134
|
-
conda create -n mlmm-toolkit python=3.
|
|
135
|
+
conda create -n mlmm-toolkit python=3.12 -y && conda activate mlmm-toolkit
|
|
135
136
|
conda install -c conda-forge ambertools pdbfixer -y
|
|
136
|
-
pip install torch==2.
|
|
137
|
+
pip install torch==2.13.0 --index-url https://download.pytorch.org/whl/cu130
|
|
137
138
|
|
|
138
|
-
# 2.
|
|
139
|
-
pip install -
|
|
139
|
+
# 2. Install mlmm-toolkit
|
|
140
|
+
pip install mlmm-toolkit
|
|
140
141
|
|
|
141
142
|
# 3. Authenticate Hugging Face once (only required for the default UMA backend)
|
|
142
143
|
# Accept the FAIR Chemistry License v1 at https://huggingface.co/facebook/UMA, then:
|
|
@@ -48,18 +48,18 @@ The source repository includes full-system COMT and [BezA](examples/beza/README.
|
|
|
48
48
|
| GPU / CUDA / VRAM | A backend-compatible NVIDIA GPU/driver for GPU execution; size VRAM from a representative target-system pilot. |
|
|
49
49
|
| RAM / Disk | Size RAM and disk for the selected backend, model cache, topology tools, and expected artifacts. |
|
|
50
50
|
|
|
51
|
-
|
|
51
|
+
**AmberTools** (`tleap`) is required to generate a topology with `mm-parm` or `all`; reuse an existing matching `--parm` to skip that preparation. The default `hessian_ff` backend needs **a C++20-capable compiler** (validated with GCC 13.3) to JIT-compile native kernels on first use. **pdbfixer** is optional — only `mm-parm --add-h` needs it — `conda install -c conda-forge ambertools pdbfixer -y` installs both. CPU-only ML/MM execution is supported but can be substantially slower than GPU execution; benchmark the selected backend and system. Full requirement and tuning details: [docs/getting-started.md#installation](docs/getting-started.md#installation).
|
|
52
52
|
|
|
53
53
|
## Installation
|
|
54
54
|
|
|
55
55
|
```bash
|
|
56
56
|
# 1. New env + AmberTools + CUDA-enabled PyTorch
|
|
57
|
-
conda create -n mlmm-toolkit python=3.
|
|
57
|
+
conda create -n mlmm-toolkit python=3.12 -y && conda activate mlmm-toolkit
|
|
58
58
|
conda install -c conda-forge ambertools pdbfixer -y
|
|
59
|
-
pip install torch==2.
|
|
59
|
+
pip install torch==2.13.0 --index-url https://download.pytorch.org/whl/cu130
|
|
60
60
|
|
|
61
|
-
# 2.
|
|
62
|
-
pip install -
|
|
61
|
+
# 2. Install mlmm-toolkit
|
|
62
|
+
pip install mlmm-toolkit
|
|
63
63
|
|
|
64
64
|
# 3. Authenticate Hugging Face once (only required for the default UMA backend)
|
|
65
65
|
# Accept the FAIR Chemistry License v1 at https://huggingface.co/facebook/UMA, then:
|
|
@@ -27,7 +27,7 @@ PDB and public CIF companions restore the original identifiers.
|
|
|
27
27
|
terminal saddle order separately. `all` proceeds to IRC only when optimization
|
|
28
28
|
converged, terminal PHVA completed, and a negative reaction direction is
|
|
29
29
|
available. A converged higher-order stationary point may continue through
|
|
30
|
-
warning-
|
|
30
|
+
warning-labeled **diagnostic** IRC, but it is not a certified first-order TS.
|
|
31
31
|
Actual optimizer non-convergence, zero imaginary modes, failed/unavailable PHVA,
|
|
32
32
|
or no valid negative root stops after preserving TS artifacts and before IRC.
|
|
33
33
|
`--skip-final-freq` also stops before IRC because the reaction direction cannot
|
|
@@ -95,19 +95,19 @@ artifact and is always written for PDB input.
|
|
|
95
95
|
- Omitting `-c/--center` skips extraction and uses the full input structures directly.
|
|
96
96
|
2. **ML/MM preparation (parm7 + layer assignment)**
|
|
97
97
|
- `mm_parm` runs once on the first full input PDB and writes `<out-dir>/mm_parm/<input_basename>.parm7` / `.rst7` (a reusable deliverable you can pass back as `--parm`), which are passed automatically as `--parm`.
|
|
98
|
-
- `define-layer` runs on each full-system PDB and assigns 3-layer B-factors (ML = 0.0, Movable-MM = 10.0, Frozen = 20.0) based on the ML-region definition. The layered full-system PDBs are written under `<out-dir>/layered/`.
|
|
98
|
+
- `define-layer` runs on each full-system PDB and assigns 3-layer B-factors (ML = 0.0, Movable-MM = 10.0, Frozen-MM = 20.0) based on the ML-region definition. The layered full-system PDBs are written under `<out-dir>/layered/`.
|
|
99
99
|
3. **Optional staged scan** (single-structure only)
|
|
100
100
|
- When exactly one input PDB is provided and `--scan-lists` is given, the tool performs a staged, bond-length-driven scan on the layered full-system PDB using the ML/MM calculator.
|
|
101
101
|
- Each stage's relaxed structure (`stage_XX/result.pdb`) is collected as an intermediate / product candidate. The ordered input series for the path search becomes `[initial layered PDB, stage_01/result.pdb, stage_02/result.pdb, ...]`.
|
|
102
102
|
4. **MEP search on full-system layered PDBs**
|
|
103
103
|
- All MEP calculations run on full-system layered PDBs (with `--parm` and `--detect-layer`), not on pockets.
|
|
104
|
-
- **`--refine-path`** runs recursive `path_search` with automatic refinement,
|
|
104
|
+
- **`--refine-path`** runs recursive `path_search` with automatic refinement, proposing multistep reaction-path candidates for TS/IRC validation. It also refines a single-step MEP, relaxing the HEI region on a path that needs no subdivision. `--max-depth` caps the subdivision levels (`0` disables subdivision). Complex multistep mechanisms may need manual trial-and-error to obtain a converged pathway.
|
|
105
105
|
- Select GSM (default) or DMF with `--mep-mode`. `--dmf-backend gpu` uses `dmf.torch`; use `--dmf-backend cpu` for the NumPy implementation or after a GPU out-of-memory error.
|
|
106
106
|
- **`--no-refine-path` (default)** runs `path-opt` with the selected optimizer per adjacent pair, then concatenates trajectories, extracts the HEI per segment, detects bond changes, and writes `summary.json`. Both modes support Stage 5 post-processing.
|
|
107
107
|
- For multi-input runs, the original full PDBs are supplied as merge references automatically. In the scan-derived series (single-structure case), the single original full PDB is reused as the reference template.
|
|
108
108
|
5. **Summary and optional post-processing**
|
|
109
109
|
- The raw MEP-engine output (per-segment trajectories, the full MEP trajectory, and the engine `summary.json`) is written under `<out-dir>/_work/path_opt/` (or `<out-dir>/_work/path_search/` with `--refine-path`); the merged products (`mep.pdb`, optional `mep.cif`, `mep_trj.xyz`, `mep_plot.png`, `energy_diagram_MEP.png`) are moved to `<out-dir>/` and `summary.{json,log}` copied there.
|
|
110
|
-
- `--tsopt` runs TS optimization on each HEI. After the TS gate, `all` continues with EulerPC IRC and segment energy diagrams.
|
|
110
|
+
- `--tsopt` runs TS optimization on each HEI. After the TS gate, `all` continues with EulerPC IRC and segment energy diagrams. Both `--thermo` and `--dft` require `--tsopt`.
|
|
111
111
|
- `--thermo` computes ML/MM thermochemistry on (R, TS, P) and adds a Gibbs diagram.
|
|
112
112
|
- `--dft` runs model-region DFT single-points on (R, TS, P) and adds a model-DFT electronic diagram. With `--thermo`, the subtractive DFT//MLIP/MM total plus the ML/MM thermal correction produces the DFT//MLIP/MM Gibbs diagram.
|
|
113
113
|
- TS optimization, IRC, frequency analysis, and flatten PHVA use the fixed constrained treatment, which removes only full-system rigid motions that leave frozen anchors fixed; realistic ML/MM boundaries normally have effective rank 0.
|
|
@@ -116,6 +116,12 @@ artifact and is always written for PDB input.
|
|
|
116
116
|
- Skips the MEP search and runs `tsopt` on the layered full-system PDB. After the TS gate, it performs EulerPC IRC, minimizes both ends, and optionally adds thermochemistry, DFT, and DFT//MLIP/MM diagrams.
|
|
117
117
|
- When IRC runs, its ends are emitted as chemically unassigned `E1` and `E2` because no path/reference orientation is available. The summary reports the barrier from each endpoint to TS and does not emit R/P reaction energies. Inspect the structures before assigning chemical identities.
|
|
118
118
|
|
|
119
|
+
An endpoint execution error or missing valid final structure stops that segment
|
|
120
|
+
before frequency/DFT and refined diagrams. `endpoint_opt/failure.json` records
|
|
121
|
+
the error; TS/IRC structures and endpoint diagnostics are retained. Diagnostic
|
|
122
|
+
frequency/DFT calculations may still run after ordinary nonconvergence with
|
|
123
|
+
finite output.
|
|
124
|
+
|
|
119
125
|
## Outputs
|
|
120
126
|
|
|
121
127
|
The tree has three zones: **deliverables at the root**, **per-segment deliverables under `segments/seg_NN/`**, and **pipeline scratch under `_work/`** (safe to remove once you have the results). The three you check first are `summary.log`, `summary.json`, and `mep.pdb` (the concatenated reaction path, moved to the root; raw engine output stays under `_work/path_opt/` by default, or `_work/path_search/` with `--refine-path`).
|
|
@@ -160,7 +166,7 @@ The tree has three zones: **deliverables at the root**, **per-segment deliverabl
|
|
|
160
166
|
|
|
161
167
|
In **TSOPT-only mode** (single input + `--tsopt`, no `--scan-lists`) there is no MEP stage. `ts/` is written under `segments/seg_01/`; after the TS gate, the E1/TS/E2 structures and `irc/` are added there, followed by requested `freq/` and `dft/` outputs. `_work/path_opt/` is absent.
|
|
162
168
|
|
|
163
|
-
At `-v 2` the console
|
|
169
|
+
At `-v 2` the console summarizes extraction, MM preparation, scan stages, MEP progress, and per-stage timing; see {ref}`verbosity-levels`.
|
|
164
170
|
|
|
165
171
|
### Reading `summary.log`
|
|
166
172
|
|
|
@@ -180,6 +186,8 @@ The log is organized into numbered sections:
|
|
|
180
186
|
|
|
181
187
|
### Reading `summary.json`
|
|
182
188
|
|
|
189
|
+
Before interpreting energies, check `scientific_status` and `scientific_status_reasons`; see [result status and stage outcomes](json-output.md#execution-and-scientific-truth) for partial or stopped runs.
|
|
190
|
+
|
|
183
191
|
Top-level keys: `out_dir`, `n_images`, `n_segments` (run metadata and counts); `segments` (per-segment entries with `index`, `tag`, `kind`, `barrier_kcal`, `delta_kcal`, `bond_changes`); `energy_diagrams` (optional payloads with `labels`, `energies_kcal`, `energies_au`, `ylabel`, `image` paths).
|
|
184
192
|
|
|
185
193
|
When stage `result.json` files or `thermoanalysis.yaml` are written, their
|
|
@@ -235,9 +243,9 @@ Defaults shown are used when the option is not specified. The full flag list is
|
|
|
235
243
|
### MEP search
|
|
236
244
|
|
|
237
245
|
```{note}
|
|
238
|
-
`--max-cycles-gsm` and `--max-cycles-dmf` control only the selected MEP
|
|
239
|
-
|
|
240
|
-
|
|
246
|
+
`--max-cycles-gsm` and `--max-cycles-dmf` control only the selected MEP stage.
|
|
247
|
+
Scan, TS optimization, IRC, and other stages keep their dedicated cycle options
|
|
248
|
+
and defaults.
|
|
241
249
|
```
|
|
242
250
|
|
|
243
251
|
| Option | Description | Default |
|
|
@@ -246,18 +254,19 @@ their dedicated cycle options and defaults.
|
|
|
246
254
|
| `--mep-mode [gsm\|dmf]` | MEP optimizer forwarded to both `path-opt` and recursive `path-search`. | `gsm` |
|
|
247
255
|
| `--dmf-backend [gpu\|cpu]` | DMF implementation. The parent forwards this only when explicitly set, so a child YAML `dmf.backend` remains effective otherwise. | `gpu` |
|
|
248
256
|
| `--max-nodes INT` | Internal nodes per GSM/DMF segment. | `20` |
|
|
257
|
+
| `--max-depth INT` | Recursive subdivision levels allowed; requires `--refine-path`. `0` disables subdivision, returning each input pair as one MEP segment (none when its HEI sits at an endpoint). A capped interval is tagged `seg_NNN_maxdepth` and may hold more than one step. | `10` |
|
|
249
258
|
| `--gsm-param [equi\|energy]` | GSM node parameterization after string growth. `energy` concentrates nodes in high-energy regions and may be tried when an equidistant path skips the reaction-coordinate region near the HEI; it does not identify a TS. | `equi` |
|
|
250
259
|
| `--max-cycles-gsm INT` | GSM string-optimizer cycle cap for the MEP child. | `300` |
|
|
251
|
-
| `--max-cycles-dmf INT` | DMF IPOPT iteration cap for the MEP child. | `
|
|
260
|
+
| `--max-cycles-dmf INT` | DMF IPOPT iteration cap for the MEP child. | `3000` |
|
|
252
261
|
| `--climb / --no-climb` | Enable climbing-image TS refinement where supported by the selected optimizer. | `True` |
|
|
253
262
|
| `--opt-mode [grad\|hess]` | Fallback preset for TSOPT and post-IRC endpoint optimization (`grad` → Dimer / L-BFGS, `hess` → RS-P-RFO / RFO). `--opt-mode-post` takes precedence. | `grad` |
|
|
254
263
|
| `--opt-mode-post [grad\|hess]` | Optimizer preset override for TSOPT / post-IRC endpoint optimizations (`grad` → Dimer / L-BFGS, `hess` → RS-P-RFO / RFO). | `hess` |
|
|
255
264
|
| `--thresh TEXT` | Convergence preset for single-structure optimizations and scan relaxations (`gau_loose`, `gau`, `gau_tight`, `gau_vtight`, `baker`, `never`). | `gau` |
|
|
256
265
|
| `--thresh-gsm TEXT` | Convergence preset for the GSM string optimizer of the MEP stage (same presets as `--thresh`). | `gau_loose` |
|
|
257
266
|
| `--thresh-dmf TEXT` | IPOPT dual-infeasibility tolerance of the DMF MEP stage: `tight` (0.04), `middle` (0.10), `loose` (0.20), or a positive float. Not a Gaussian preset. | `tight` |
|
|
258
|
-
| `--thresh-post TEXT` | Convergence preset for post-IRC endpoint optimizations. | `baker` |
|
|
267
|
+
| `--thresh-post TEXT` | Convergence preset for TS and post-IRC endpoint optimizations. | `baker` |
|
|
259
268
|
| `--preopt / --no-preopt` | Pre-optimize endpoints before segmentation. | `True` |
|
|
260
|
-
| `--refine-path / --no-refine-path` | `--no-refine-path` (default) → single-pass `path-opt`; `--refine-path` → recursive `path-search
|
|
269
|
+
| `--refine-path / --no-refine-path` | `--no-refine-path` (default) → single-pass `path-opt`; `--refine-path` → recursive `path-search`, which proposes multistep reaction-path candidates and also refines a single-step MEP, where it can improve a poor HEI or TS estimate. Both modes support Stage 5 (TSOPT / thermo / DFT). | `False` |
|
|
261
270
|
| `-b, --backend CHOICE` | MLIP backend for the ML region: `uma` (default), `orb`, `mace`, `aimnet2`. | `uma` |
|
|
262
271
|
| `--precision [fp32\|fp64]` | Backend precision. Unset uses UMA/AIMNet2 fp32 and ORB/MACE fp64. AIMNet2 rejects fp64. | backend-specific |
|
|
263
272
|
| `--workers INT` | UMA predictor workers. Values greater than 1 require `fairchem-core[extras]` and are incompatible with an analytical Hessian. | `1` |
|
|
@@ -62,7 +62,7 @@ name from the CLI.
|
|
|
62
62
|
### UMA fp64
|
|
63
63
|
|
|
64
64
|
Switching OMol-trained UMA from the default fp32 to fp64 can have a non-trivial
|
|
65
|
-
impact on
|
|
65
|
+
impact on TSOPT and Hessian evaluation. For full-system PDBs with valid B-factor layers:
|
|
66
66
|
|
|
67
67
|
```bash
|
|
68
68
|
mlmm tsopt -i ts.pdb --parm real.parm7 -q 0 -m 1 --precision fp64
|
|
@@ -125,7 +125,7 @@ def get_calculator(charge=0, spin=1, device="auto", **kwargs):
|
|
|
125
125
|
Swap `EMT()` for the engine you want — e.g. `tblite.ase.TBLite(...)` for
|
|
126
126
|
GFN-xTB, the DFTB+ ASE calculator, or `ase.calculators.orca.ORCA(...)`. Then
|
|
127
127
|
pass the file to a stage or to `all` (it selects the `custom` ML backend,
|
|
128
|
-
overriding `--backend`):
|
|
128
|
+
overriding `--backend`). These examples require full-system PDBs with valid B-factor layers:
|
|
129
129
|
|
|
130
130
|
mlmm sp -i complex.pdb --parm system.parm7 --calc-file my_calc.py -q 0 -m 1
|
|
131
131
|
mlmm opt -i complex.pdb --parm system.parm7 --calc-file my_calc.py -q 0 -m 1
|
|
@@ -164,7 +164,7 @@ To add a new backend `XYZModel` exposed as `--backend xyz`:
|
|
|
164
164
|
eV/Å, plus a backend-specific opaque object), `hessian_analytical(opaque, n_atoms,
|
|
165
165
|
*, dtype) -> torch.Tensor` (returns the Hessian in eV/Ų), and the
|
|
166
166
|
`supports_analytical_hessian` and `device` properties. Inherit from `_MLBackend`
|
|
167
|
-
to
|
|
167
|
+
to inherit the generic finite-difference `hessian_fd(...)` (used when the
|
|
168
168
|
backend has no analytical Hessian).
|
|
169
169
|
3. **Register in `_create_ml_backend`**: extend the factory in
|
|
170
170
|
`mlmm/backends/mlmm_calc.py` to dispatch `backend == "xyz"` to `_XYZBackend(...)`.
|
|
@@ -4,25 +4,18 @@ Conventions shared across all `mlmm` commands.
|
|
|
4
4
|
|
|
5
5
|
## Boolean options
|
|
6
6
|
|
|
7
|
-
|
|
7
|
+
Use paired flags for on/off options:
|
|
8
8
|
|
|
9
9
|
| Form | Example |
|
|
10
10
|
|---|---|
|
|
11
11
|
| Positive flag | `--tsopt` |
|
|
12
12
|
| Negative flag | `--no-tsopt` |
|
|
13
|
-
| Positive value | `--tsopt True`, `--tsopt yes`, `--tsopt 1`, `--tsopt on` |
|
|
14
|
-
| Negative value | `--tsopt False`, `--tsopt no`, `--tsopt 0`, `--tsopt off` |
|
|
15
13
|
|
|
16
14
|
```bash
|
|
17
15
|
--tsopt --thermo --no-dft # toggle form
|
|
18
|
-
--tsopt True --thermo yes --dft 0 # value form
|
|
19
|
-
--tsopt true --thermo on --dft off # mix is fine
|
|
20
16
|
```
|
|
21
17
|
|
|
22
|
-
|
|
23
|
-
|
|
24
|
-
`--detect-layer` is different: layer detection is automatic and enabled by
|
|
25
|
-
default, so it is a positive-only marker rather than an on/off toggle.
|
|
18
|
+
Use paired flags (`--tsopt` / `--no-tsopt`) for new commands. Value forms remain accepted for compatibility.
|
|
26
19
|
|
|
27
20
|
Common toggles: `--tsopt` / `--thermo` / `--dft` (post-processing stages) · `--dump` (write trajectory files) · `--preopt` / `--endopt` (pre/post optimization) · `--climb` (climbing-image MEP).
|
|
28
21
|
|
|
@@ -93,7 +86,7 @@ mlmm opt -i input.pdb --parm real.parm7 -q -1 --show-config --dry-run
|
|
|
93
86
|
|---|---|---|
|
|
94
87
|
| ML | 0.0 | MLIP energy / force / Hessian |
|
|
95
88
|
| Movable-MM | 10.0 | MM atoms free to move |
|
|
96
|
-
| Frozen | 20.0 | Coordinates fixed (non-bonded interactions still included) |
|
|
89
|
+
| Frozen-MM | 20.0 | Coordinates fixed (non-bonded interactions still included) |
|
|
97
90
|
|
|
98
91
|
Tolerance ±1.0 when reading B-factors. Inspect visually by coloring on B-factor. Four ways to assign layers:
|
|
99
92
|
|
|
@@ -106,7 +99,7 @@ mlmm define-layer -i system.pdb --model-pdb ml_region.pdb -o labeled.pdb
|
|
|
106
99
|
# 2. Distance cutoffs (YAML)
|
|
107
100
|
calc:
|
|
108
101
|
hess_cutoff: 3.6 # Hessian-target MM
|
|
109
|
-
movable_cutoff: 8.0 # Movable-MM (beyond → Frozen)
|
|
102
|
+
movable_cutoff: 8.0 # Movable-MM (beyond → Frozen-MM)
|
|
110
103
|
|
|
111
104
|
# 3. Read existing B-factors
|
|
112
105
|
calc:
|
|
@@ -15,9 +15,9 @@ Full system(s) (PDB/mmCIF/XYZ)
|
|
|
15
15
|
| |
|
|
16
16
|
| ML region structure(s) (PDB)
|
|
17
17
|
| |
|
|
18
|
-
| +- Amber topology [mm-parm] <- generates parm7/rst7 via AmberTools
|
|
18
|
+
| +- Amber topology [mm-parm, full system] <- generates parm7/rst7 via AmberTools
|
|
19
19
|
| | |
|
|
20
|
-
| +- 3-layer assignment [define-layer] <- B-factor encoding for ML/MM layers
|
|
20
|
+
| +- 3-layer assignment [define-layer, full system + ML selection] <- B-factor encoding for ML/MM layers
|
|
21
21
|
| | |
|
|
22
22
|
| | v
|
|
23
23
|
| | ML/MM system (PDB with B-factors)
|
|
@@ -52,14 +52,14 @@ A central concept in mlmm-toolkit is the **3-layer ML/MM partitioning** of the s
|
|
|
52
52
|
|-------|----------|-------------|
|
|
53
53
|
| **ML** (Layer 1) | 0.0 | The reactive region. Full MLIP energy, forces, and Hessian. |
|
|
54
54
|
| **Movable-MM** (Layer 2) | 10.0 | MM atoms allowed to move during optimization. |
|
|
55
|
-
| **Frozen** (Layer 3) | 20.0 | Coordinates are fixed; no optimization. |
|
|
55
|
+
| **Frozen-MM** (Layer 3) | 20.0 | Coordinates are fixed; no optimization. |
|
|
56
56
|
|
|
57
57
|
B-factor values are encoded in PDB columns 61-66 (the temperature factor column).
|
|
58
58
|
|
|
59
59
|
The [`define-layer`](define-layer.md) subcommand assigns these B-factors based on distance from the ML region:
|
|
60
60
|
|
|
61
61
|
- Atoms/residues within `--radius-freeze` (default 8.0 Å) are assigned to Movable-MM.
|
|
62
|
-
- Atoms/residues beyond `--radius-freeze` are Frozen.
|
|
62
|
+
- Atoms/residues beyond `--radius-freeze` are Frozen-MM.
|
|
63
63
|
|
|
64
64
|
Hessian-target MM atoms are controlled by calculator options (`hess_cutoff`, explicit `hess_mm_atoms`, etc.), not by a dedicated B-factor layer.
|
|
65
65
|
|
|
@@ -242,7 +242,7 @@ There are two ways to define the ML region:
|
|
|
242
242
|
set the model charge explicitly with `-q`. This is the safer choice when you have
|
|
243
243
|
hand-edited atoms (custom truncation, protonation / charge changes) that automatic
|
|
244
244
|
derivation cannot infer, or when the topology is too unusual for automatic
|
|
245
|
-
extraction. `--modified-residue`
|
|
245
|
+
extraction. `--modified-residue` does not apply on this path. You do **not**
|
|
246
246
|
need to add link hydrogens to the model PDB — the ML/MM calculator inserts them from
|
|
247
247
|
the `--parm` topology at the ML/MM boundary.
|
|
248
248
|
|
|
@@ -267,6 +267,7 @@ calc:
|
|
|
267
267
|
- ["A:LIG:1:C1", "A:ALA:2:CA"]
|
|
268
268
|
```
|
|
269
269
|
|
|
270
|
+
(model-pdb-selection)=
|
|
270
271
|
#### How to construct a reliable `model.pdb`
|
|
271
272
|
|
|
272
273
|
`model.pdb` is an **atom-selection file**, not an independently rebuilt cluster.
|
|
@@ -119,16 +119,15 @@ hostname
|
|
|
119
119
|
cd "${PBS_O_WORKDIR}"
|
|
120
120
|
|
|
121
121
|
# hessian_ff JIT-compiles C++ kernels on first use. If the system compiler is
|
|
122
|
-
# missing or
|
|
122
|
+
# missing or cannot compile in C++20 mode, load the site's compiler module here:
|
|
123
123
|
# module load <COMPILER_MODULE>
|
|
124
124
|
|
|
125
125
|
# Activate conda environment
|
|
126
126
|
source ~/miniconda3/etc/profile.d/conda.sh
|
|
127
127
|
conda activate <your-env>
|
|
128
128
|
command -v g++ >/dev/null || { echo "g++ is required for hessian_ff" >&2; exit 1; }
|
|
129
|
-
|
|
130
|
-
|
|
131
|
-
echo "hessian_ff requires GCC >= 9 (found major: $gxx_major)" >&2
|
|
129
|
+
if ! g++ -std=c++20 -x c++ -fsyntax-only /dev/null; then
|
|
130
|
+
echo "hessian_ff requires a compiler supporting PyTorch's C++20 JIT flag" >&2
|
|
132
131
|
exit 1
|
|
133
132
|
fi
|
|
134
133
|
command -v ninja >/dev/null || { echo "ninja is required for hessian_ff" >&2; exit 1; }
|
|
@@ -163,9 +162,8 @@ hostname
|
|
|
163
162
|
source ~/miniconda3/etc/profile.d/conda.sh
|
|
164
163
|
conda activate <your-env>
|
|
165
164
|
command -v g++ >/dev/null || { echo "g++ is required for hessian_ff" >&2; exit 1; }
|
|
166
|
-
|
|
167
|
-
|
|
168
|
-
echo "hessian_ff requires GCC >= 9 (found major: $gxx_major)" >&2
|
|
165
|
+
if ! g++ -std=c++20 -x c++ -fsyntax-only /dev/null; then
|
|
166
|
+
echo "hessian_ff requires a compiler supporting PyTorch's C++20 JIT flag" >&2
|
|
169
167
|
exit 1
|
|
170
168
|
fi
|
|
171
169
|
command -v ninja >/dev/null || { echo "ninja is required for hessian_ff" >&2; exit 1; }
|
|
@@ -184,7 +182,7 @@ mlmm opt \
|
|
|
184
182
|
- **CPU threads:** Request enough CPUs for the MM backend (`mm_threads`, default 16). Set `ppn=32` (PBS) or `--cpus-per-task=32` (Slurm) for a safety margin.
|
|
185
183
|
- **Memory:** size RAM from a representative pilot and scheduler peak-memory logs.
|
|
186
184
|
- **CUDA runtime:** Official PyTorch wheels carry CUDA user-space libraries; a compatible NVIDIA driver is normally sufficient. Load a site CUDA toolkit only for an extension that needs it.
|
|
187
|
-
- **C++ compiler:** The default `hessian_ff` MM backend JIT-compiles C++ kernels on first use, independently of CUDA. Every compute node needs
|
|
185
|
+
- **C++ compiler:** The default `hessian_ff` MM backend JIT-compiles C++ kernels on first use, independently of CUDA. Every compute node needs a C++20-capable compiler and Ninja (GCC 13.3 was validated); load a compiler module when the system `g++` is absent or too old.
|
|
188
186
|
|
|
189
187
|
### Specifying a GPU index
|
|
190
188
|
|
|
@@ -152,12 +152,14 @@ Register each unlisted name with its integer charge, for example
|
|
|
152
152
|
`--modified-residue HD1:0,HE1:0,CM1:0,AP1:0`. A residue already in the
|
|
153
153
|
catalog may omit `:charge` and keeps its catalog charge (for example, `SEP`
|
|
154
154
|
remains −2). `extract` then treats the names as amino acids and applies
|
|
155
|
-
backbone truncation
|
|
155
|
+
backbone truncation and charge assignment
|
|
156
156
|
automatically, and the warning above is suppressed.
|
|
157
157
|
```
|
|
158
158
|
|
|
159
159
|
```{important}
|
|
160
|
-
If `--modified-residue` cannot cover your case (e.g. unusual backbone topology),
|
|
160
|
+
If `--modified-residue` cannot cover your case (e.g. unusual backbone topology), follow the {ref}`manual ML-selection recipe <model-pdb-selection>` to prepare `--model-pdb` without adding atoms.
|
|
161
|
+
|
|
162
|
+
For a **standalone capped pocket**, not an ML-selection file:
|
|
161
163
|
|
|
162
164
|
1. Select residues around the active site and determine truncation points.
|
|
163
165
|
2. Add a link hydrogen on the parent atom (the atom that remains) of each severed covalent bond.
|
|
@@ -1,5 +1,10 @@
|
|
|
1
1
|
# `freq`
|
|
2
2
|
|
|
3
|
+
`frequencies_cm-1.txt`, JSON `frequencies_cm`, and `n_modes` retain the complete signed physical spectrum after the existing frozen-atom and rigid-mode projection. `--max-write` and `--sort` control only which mode files are written and their order. `n_imaginary` (YAML `num_imag_freq`) is the resolved negative count below the reporting threshold; `n_negative_modes` also includes weak negative modes. With `frequency_representation: complete`, `near_zero_frequencies_cm` is a subset of the complete array; do not append it and count modes twice.
|
|
4
|
+
|
|
5
|
+
Thermochemistry retains the existing QRRHO policy (100 cm⁻¹ rotor cutoff, no imaginary inversion and no positive-frequency floor), including positive low-frequency modes. Changing `freq.zero_cutoff_cm` does not change thermal values computed from the same complete spectrum.
|
|
6
|
+
|
|
7
|
+
|
|
3
8
|
Compute ML/MM vibrational frequencies and thermochemistry (zero-point energy (ZPE), Gibbs energy, etc.) on a layered enzyme PDB, with partial-Hessian vibrational analysis (PHVA) support.
|
|
4
9
|
|
|
5
10
|
**When to use `mlmm freq`:**
|
|
@@ -15,6 +20,8 @@ representative pilot.
|
|
|
15
20
|
|
|
16
21
|
## Examples
|
|
17
22
|
|
|
23
|
+
Here, `pocket.pdb` contains the full system matching `real.parm7`; `ml_region.pdb` selects the ML atoms.
|
|
24
|
+
|
|
18
25
|
Basic frequency analysis:
|
|
19
26
|
|
|
20
27
|
```bash
|
|
@@ -176,7 +183,7 @@ calc:
|
|
|
176
183
|
mm_fd: true # MM finite-difference toggle
|
|
177
184
|
return_partial_hessian: true # allow partial Hessians (PHVA default)
|
|
178
185
|
freq:
|
|
179
|
-
zero_cutoff_cm: 5.0 #
|
|
186
|
+
zero_cutoff_cm: 5.0 # Near-zero classification window (cm^-1); modes are retained
|
|
180
187
|
amplitude_ang: 0.8 # displacement amplitude for modes (Å)
|
|
181
188
|
n_frames: 20 # number of frames per mode
|
|
182
189
|
max_write: 10 # maximum number of modes to write
|