Cheminformatics

Organization

awslabs

cheminformatics

Cheminformatics pipeline for small-molecule property calculation, filtering, and similarity analysis using RDKit. Use when the user asks to compute molecular descriptors, filter compounds by Lipinski or Veber rules, detect PAINS, calculate fingerprint similarity, run matched molecular pair analysis, generate ADMET descriptors, or process SMILES. Triggers include "RDKit", "molecular descriptors", "Lipinski", "rule of five", "Veber", "PAINS", "pan-assay interference", "Morgan fingerprint", "Tanimoto", "fingerprint similarity", "matched molecular pair", "MMP", "mmpdb", "ADMET", "druglikeness", "SMILES", "cheminformatics", "compound filtering", "chemical similarity".

Overview

Publisher	awslabs
Repository	`hcls-agent-skills`
Skill name	`cheminformatics`
Stars	71
Forks	14
Bundled files	Instructions only
License	MIT-0
Links	Skill source Repository Homepage

Markdown instructions
A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.
Works with any LLM
AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.
Self-contained
Everything the model needs lives in the instructions — no extra files to sync.
Open source
Published by awslabs on GitHub. Read the source before you install it.

Installation

Install the Cheminformatics AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

Open the app and go to Plugins → Skills.
Choose "Install from GitHub".
Paste the skill folder URL below and confirm.
https://github.com/awslabs/hcls-agent-skills/tree/main/skills/cheminformatics
Enable the skill in any chat where you want it available.

Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills

git clone --depth 1 https://github.com/awslabs/hcls-agent-skills.git /tmp/hcls-agent-skills
mkdir -p .claude/skills
cp -r /tmp/hcls-agent-skills/skills/cheminformatics .claude/skills/cheminformatics

Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Cheminformatics in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Cheminformatics on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Cheminformatics is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

Loaded only when it is needed
The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.
Switch models mid-chat
Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Cheminformatics (RDKit) — Pipeline Skill (Thin Scaffold)

Overview

Adds fingerprint selection logic, filter cascade ordering, and version-specific gotchas for RDKit workflows. The agent already knows RDKit syntax — this skill prevents parameter mismatches and deprecated API usage.

Usage

Activate when filtering compound libraries (Lipinski/Veber/PAINS)
Activate for fingerprint similarity search or clustering
Activate for matched molecular pair (MMP) analysis with mmpdb

Core Concepts

Filter cascade order: Lipinski → Veber → PAINS → similarity (cheapest first, most expensive last).

Fingerprint API: Use rdFingerprintGenerator module exclusively (since RDKit 2024.03). The AllChem.GetMorganFingerprintAsBitVect() path is deprecated and will be removed.

SMILES validation: Always parse with sanitize=False then explicit SanitizeMol() in try/except. Default MolFromSmiles silently returns None on partial failures.

Decision Logic

Fingerprint Selection

Goal	Fingerprint	Parameters	Tanimoto threshold
General similarity	Morgan (ECFP4)	radius=2, fpSize=2048	≥ 0.7
High-specificity matching	Morgan (ECFP6)	radius=3, fpSize=4096	≥ 0.6
Substructure screening	MACCS 166 keys	—	≥ 0.8
Pharmacophore similarity	FCFP	radius=2, `GetMorganFeatureAtomInvGen()`	≥ 0.6
Large-scale clustering	Morgan	radius=2, fpSize=1024	Speed over precision

Filter Thresholds

Filter	Rule	Threshold	Notes
Lipinski MW	≤ 500 Da	Allow ≤1 violation total	Not binary pass/fail
Lipinski LogP	≤ 5	—	—
Lipinski HBD	≤ 5	—	—
Lipinski HBA	≤ 10	—	—
Veber TPSA	≤ 140 Å²	Hard cutoff	—
Veber RotBonds	≤ 10	Hard cutoff	—
PAINS	A+B+C catalogs	HTS triage only	Don't apply to advanced compounds

MMP Analysis Decision

Scenario	Tool	Key parameter
SAR from compound pairs	mmpdb fragment → index → transform	`--min-pairs 3`
Property prediction	mmpdb predict	Requires indexed DB with property data
Core identification	`rdFMCS.FindMCS()`	`ringMatchesRingOnly=True, completeRingsOnly=True`
Large libraries (>10K)	mmpdb CLI	`--max-variable-size 10` to limit fragmentation

Critical Parameters

Parameter	Value	When to change
Morgan radius	2 (ECFP4)	3 for higher specificity
fpSize	2048	1024 for speed; 4096 for precision
includeChirality	False	True for stereo-sensitive applications
Lipinski max_violations	1	0 for strict; exempt biologics entirely
mmpdb --min-pairs	3	Lower for sparse datasets
mmpdb --max-variable-size	10	Increase for macrocycle transformations
PAINS catalogs	A+B+C (all three)	Never use partial sets

Common Mistakes

Wrong: AllChem.GetMorganFingerprintAsBitVect() → Right: rdFingerprintGenerator.GetMorganGenerator(radius=2, fpSize=2048).GetFingerprint(mol) → Why: AllChem FP functions deprecated since RDKit 2024.03, will be removed
Wrong: Default MolFromSmiles(smi) without error handling → Right: MolFromSmiles(smi, sanitize=False) + SanitizeMol() in try/except → Why: Default silently returns None; explicit sanitize catches partial failures
Wrong: Lipinski as hard binary filter → Right: Allow ≤1 violation → Why: Many approved drugs violate one rule; binary filtering rejects valid candidates
Wrong: PAINS on clinical candidates → Right: PAINS only for HTS library triage → Why: Designed for screening artifacts; flags valid advanced compounds
Wrong: Morgan radius=1 → Right: radius=2 minimum (ECFP4) → Why: Radius 1 loses structural information; too many false-positive similarities
Wrong: Comparing FPs with different radius or nBits → Right: Identical parameters for both FPs → Why: Tanimoto only meaningful in same parameter space
Wrong: Ignoring stereochemistry for chiral compounds → Right: includeChirality=True in generator → Why: Default treats enantiomers as identical
Wrong: Fixed 0.7 threshold for all FP types → Right: Validate threshold per fingerprint type and chemical series → Why: 0.7 convention is ECFP4-specific; other types need different cutoffs
Wrong: Raw SMILES into mmpdb → Right: Pre-canonicalize with MolToSmiles(MolFromSmiles(smi)) → Why: mmpdb expects canonical SMILES; non-canonical causes fragmentation errors
Wrong: Confusing LogP with LogD → Right: Use LogD at pH 7.4 for ionizable compounds; RDKit computes LogP only → Why: LogP overestimates lipophilicity for ionizable molecules at physiological pH

Response Format

Lead with the command or code the user needs — explain after
Structure as: confirm inputs → working code → key parameters explained → gotchas
One complete working example per task; do not show every alternative
Keep code comments minimal and functional
Target: 50-100 lines of code with brief surrounding explanation

Frequently asked questions

What does the Cheminformatics AI skill do?

Why use Cheminformatics on TypingMind?

Because you install it once and use it with any model. Cheminformatics is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Cheminformatics in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/awslabs/hcls-agent-skills/tree/main/skills/cheminformatics. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Cheminformatics?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Cheminformatics?

As many as you like. As long as a model supports skills, you can use Cheminformatics with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Cheminformatics AI skill free?

Yes. It is published on GitHub by awslabs under the MIT-0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Huggingface Hub

NousResearch

OrganizationPopular

HuggingFace hf CLI: search/download/upload models, datasets.

Pytorch Lightning

Clean training loops with built-in distributed support.

Troubleshoot

Investigate unexpected behavior in the current Copilot agent session by analyzing its event log. Use when the user asks why something happened, why a request was slow, why a tool was or was not used,…

Pdf

Use this skill whenever the user wants to do anything with PDF files. This includes reading or extracting text/tables from PDFs, combining or merging multiple PDFs into one, splitting PDFs apart,…

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

Cheminformatics (RDKit) — Pipeline Skill (Thin Scaffold)

Overview

Usage

Core Concepts

Decision Logic

Fingerprint Selection

Filter Thresholds

MMP Analysis Decision

Critical Parameters

Common Mistakes

Response Format

Frequently asked questions

Related AI Skills

Set up your own AI workspace now