rdkit-conf

Name: rdkit-conf
Author: jinzhezenggroup

113stars23forks

How to install rdkit-conf?

jinzhezenggroup/computational-chemistry-agent-skills/rdkit-conf

$npx -y skills add jinzhezenggroup/computational-chemistry-agent-skills --skill rdkit-conf

Installs into the current project.

Summary

A standardized CLI wrapper for RDKit 3D/2D conformer generation that samples multiple conformers per molecule (ETKDGv3, default 10), optimizes each with a force field (MMFF94s/UFF), keeps the lowest-energy conformer, automatically falls back to 2D layout on total embedding failur

Files · 1

View on GitHub

SKILL.md

1	# RDKit Conformer Generation
2
3	This skill provides practical command patterns for RDKit 3D/2D conformer generation
4	using the standardized CLI wrapper: `<skill_path>/scripts/rdkit_conf_helper.py`.
5
6	Key behaviors (important for Agents):
7
8	- The script prints environment detection (Python/RDKit/Pandas) by default.
9	- Multi-conformer sampling: embeds `--num-confs` conformers (default 10) per molecule
10	via `EmbedMultipleConfs`, optimizes each with the chosen force field, and keeps the
11	lowest-energy one. Set `--num-confs 1` to revert to single-conformer behavior.
12	- 2D fallback: if all 3D embedding attempts fail, `Compute2DCoords` is used instead
13	and a `[WARN]` line is printed to stderr for that molecule.
14	- Bad/illegal SMILES are skipped entirely and logged to `*.skipped.csv` (no crash).
15	- Molecules that fell back to 2D are additionally logged to `*.fallback.csv`.
16	- Each run ends with a summary line and absolute output paths:
17	- `[INFO] Done: <N_3d> 3D, <N_2d> 2D-fallback, <N_skip> skipped (total input: <N>)`
18	- `[RESULT] conf_sdf=/abs/path.sdf`
19	- `[RESULT] conf_xyz=/abs/path.xyz`
20	- `[RESULT] fallback_csv=/abs/path.fallback.csv` (only if any 2D fallbacks occurred)
21	- `[RESULT] skipped_csv=/abs/path.skipped.csv` (only if any SMILES were skipped)
22
23	## Quick Start
24
25	Check CLI help:
26
27	```bash
28	uv run <skill_path>/scripts/rdkit_conf_helper.py --help
29	uv run <skill_path>/scripts/rdkit_conf_helper.py conf --help
30	```
31
32	Disable environment printing (optional):
33
34	```bash
35	uv run <skill_path>/scripts/rdkit_conf_helper.py --no-env conf --smiles "CCO" --output out.sdf
36	```
37
38	## Core Tasks
39
40	### 1) Generate 3D conformers (SDF output, default)
41
42	Single SMILES:
43
44	```bash
45	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
46	--smiles "CCO" \
47	--output /tmp/CCO.sdf
48	```
49
50	Single SMILES with a custom molecule name:
51
52	```bash
53	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
54	--smiles "c1ccccc1" \
55	--name benzene \
56	--output /tmp/benzene.sdf
57	```
58
59	From CSV (default SMILES column: `smiles`):
60
61	```bash
62	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
63	--file data.csv \
64	--smiles-col smiles \
65	--output data.sdf
66	```
67
68	From CSV with a name column:
69
70	```bash
71	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
72	--file data.csv \
73	--smiles-col smiles \
74	--name-col compound_id \
75	--output data.sdf
76	```
77
78	From SMI (second token per line is used as name automatically):
79
80	```bash
81	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
82	--file molecules.smi \
83	--output molecules.sdf
84	```
85
86	### 2) Control conformer sampling count
87
88	Default (10 conformers sampled, lowest-energy kept):
89
90	```bash
91	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
92	--file data.csv --output data.sdf
93	```
94
95	Single conformer (fastest, least thorough):
96
97	```bash
98	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
99	--file data.csv --num-confs 1 --output data.sdf
100	```
101
102	Increase sampling for flexible or macrocyclic molecules:
103
104	```bash
105	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
106	--file data.csv --num-confs 50 --output data.sdf
107	```
108
109	### 3) Choose force-field minimization
110
111	MMFF94s (default, falls back to UFF if unavailable):
112
113	```bash
114	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
115	--file data.csv --ff mmff94s --output data.mmff.sdf
116	```
117
118	UFF (universal force field):
119
120	```bash
121	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
122	--file data.csv --ff uff --output data.uff.sdf
123	```
124
125	Skip force-field optimization (raw ETKDG geometry only):
126
127	```bash
128	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
129	--file data.csv --ff none --output data.etkdg_raw.sdf
130	```
131
132	### 4) XYZ output
133
134	```bash
135	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
136	--file data.csv \
137	--format xyz \
138	--output data.xyz
139	```
140
141	### 5) Tuning embedding for difficult molecules
142
143	Large or macrocyclic molecules sometimes fail standard ETKDG; try random initial coordinates:
144
145	```bash
146	uv run <skill_path>/sc

People who liked this also viewed

6 picks

Type

rdkit-conf

113stars23forks

View on GitHub

How to install rdkit-conf?

jinzhezenggroup/computational-chemistry-agent-skills/rdkit-conf

$npx -y skills add jinzhezenggroup/computational-chemistry-agent-skills --skill rdkit-conf

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# RDKit Conformer Generation
2
3	This skill provides practical command patterns for RDKit 3D/2D conformer generation
4	using the standardized CLI wrapper: `<skill_path>/scripts/rdkit_conf_helper.py`.
5
6	Key behaviors (important for Agents):
7
8	- The script prints environment detection (Python/RDKit/Pandas) by default.
9	- Multi-conformer sampling: embeds `--num-confs` conformers (default 10) per molecule
10	via `EmbedMultipleConfs`, optimizes each with the chosen force field, and keeps the
11	lowest-energy one. Set `--num-confs 1` to revert to single-conformer behavior.
12	- 2D fallback: if all 3D embedding attempts fail, `Compute2DCoords` is used instead
13	and a `[WARN]` line is printed to stderr for that molecule.
14	- Bad/illegal SMILES are skipped entirely and logged to `*.skipped.csv` (no crash).
15	- Molecules that fell back to 2D are additionally logged to `*.fallback.csv`.
16	- Each run ends with a summary line and absolute output paths:
17	- `[INFO] Done: <N_3d> 3D, <N_2d> 2D-fallback, <N_skip> skipped (total input: <N>)`
18	- `[RESULT] conf_sdf=/abs/path.sdf`
19	- `[RESULT] conf_xyz=/abs/path.xyz`
20	- `[RESULT] fallback_csv=/abs/path.fallback.csv` (only if any 2D fallbacks occurred)
21	- `[RESULT] skipped_csv=/abs/path.skipped.csv` (only if any SMILES were skipped)
22
23	## Quick Start
24
25	Check CLI help:
26
27	```bash
28	uv run <skill_path>/scripts/rdkit_conf_helper.py --help
29	uv run <skill_path>/scripts/rdkit_conf_helper.py conf --help
30	```
31
32	Disable environment printing (optional):
33
34	```bash
35	uv run <skill_path>/scripts/rdkit_conf_helper.py --no-env conf --smiles "CCO" --output out.sdf
36	```
37
38	## Core Tasks
39
40	### 1) Generate 3D conformers (SDF output, default)
41
42	Single SMILES:
43
44	```bash
45	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
46	--smiles "CCO" \
47	--output /tmp/CCO.sdf
48	```
49
50	Single SMILES with a custom molecule name:
51
52	```bash
53	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
54	--smiles "c1ccccc1" \
55	--name benzene \
56	--output /tmp/benzene.sdf
57	```
58
59	From CSV (default SMILES column: `smiles`):
60
61	```bash
62	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
63	--file data.csv \
64	--smiles-col smiles \
65	--output data.sdf
66	```
67
68	From CSV with a name column:
69
70	```bash
71	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
72	--file data.csv \
73	--smiles-col smiles \
74	--name-col compound_id \
75	--output data.sdf
76	```
77
78	From SMI (second token per line is used as name automatically):
79
80	```bash
81	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
82	--file molecules.smi \
83	--output molecules.sdf
84	```
85
86	### 2) Control conformer sampling count
87
88	Default (10 conformers sampled, lowest-energy kept):
89
90	```bash
91	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
92	--file data.csv --output data.sdf
93	```
94
95	Single conformer (fastest, least thorough):
96
97	```bash
98	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
99	--file data.csv --num-confs 1 --output data.sdf
100	```
101
102	Increase sampling for flexible or macrocyclic molecules:
103
104	```bash
105	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
106	--file data.csv --num-confs 50 --output data.sdf
107	```
108
109	### 3) Choose force-field minimization
110
111	MMFF94s (default, falls back to UFF if unavailable):
112
113	```bash
114	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
115	--file data.csv --ff mmff94s --output data.mmff.sdf
116	```
117
118	UFF (universal force field):
119
120	```bash
121	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
122	--file data.csv --ff uff --output data.uff.sdf
123	```
124
125	Skip force-field optimization (raw ETKDG geometry only):
126
127	```bash
128	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
129	--file data.csv --ff none --output data.etkdg_raw.sdf
130	```
131
132	### 4) XYZ output
133
134	```bash
135	uv run <skill_path>/scripts/rdkit_conf_helper.py conf \
136	--file data.csv \
137	--format xyz \
138	--output data.xyz
139	```
140
141	### 5) Tuning embedding for difficult molecules
142
143	Large or macrocyclic molecules sometimes fail standard ETKDG; try random initial coordinates:
144
145	```bash
146	uv run <skill_path>/sc

People who liked this also viewed

6 picks

Type