skills.rest
RamanEbrahimi
RamanEbrahimiCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

pytdc

Access Therapeutics Data Commons datasets for drug discovery predictions.

3|Updated Apr 17, 2026
Search Tags:#machine learning#drug discovery#ADME#AI-ready datasets#pharmaceutical prediction
One-click install
npx skills add https://github.com/RamanEbrahimi/raman-marketplace --skill pytdc-ramanebrahimi
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: pytdc
Source: https://github.com/RamanEbrahimi/raman-marketplace/tree/main/plugins/agentic-research/skills/scientific-skills/pytdc
Command: npx skills add https://github.com/RamanEbrahimi/raman-marketplace --skill pytdc-ramanebrahimi

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires numpy, pandas, tqdm, seaborn, scikit-learn, fuzzywuzzy, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill addresses the challenge of accessing and utilizing AI-ready drug discovery datasets efficiently, providing a streamlined solution for machine learning models in pharmaceutical prediction.

Core Features & Use Cases

  • Curated Datasets: Offers a vast collection of AI-ready drug discovery datasets covering ADME, toxicity, DTI, and more.
  • Benchmarking: Facilitates benchmarking machine learning models on standardized pharmaceutical tasks.
  • Data Splits: Provides train/test splits (scaffold, cold-split) for robust model evaluation.
  • Use Case: For a drug discovery project, use this Skill to access a benchmark dataset like 'Caco2_Wang' for ADME predictions and evaluate your model's performance.

Quick Start

To access the ADME dataset 'Caco2_Wang', run the following command:

uv tdc single_pred ADME --name Caco2_Wang

Frequently Asked Questions about pytdc

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I access AI-ready drug discovery datasets for machine learning?▼

You can access curated drug discovery datasets for machine learning by using the provided command-line interface to load specific tasks like ADME prediction, which returns data formatted as pandas DataFrames ready for model training.

What types of pharmaceutical prediction tasks are supported by these datasets?▼

Supported pharmaceutical prediction tasks include single-instance predictions like ADME and toxicity, multi-instance predictions like drug-target interactions (DTI), and generation tasks, providing standardized benchmarks for machine learning models.

How do I evaluate machine learning models on pharmaceutical data using scaffold splits?▼

To evaluate models on pharmaceutical data robustly, you can load datasets that include pre-configured train and test splits such as scaffold splits and cold-splits, ensuring rigorous model performance assessment.

Do I need scikit-learn and pandas to use these drug discovery benchmarks?▼

Yes, you need pandas and scikit-learn installed in your environment, as the datasets load into pandas DataFrames for manipulation and scikit-learn is required for standard machine learning benchmarking tasks.

What is the best way to benchmark a model for ADME predictions?▼

To benchmark a model for ADME predictions, load a specific dataset like 'Caco2_Wang' using the single prediction command, then train your scikit-learn model and evaluate its performance against the provided test split.

Can I use this for drug-target interaction prediction or only single-instance ADME tasks?▼

You can use this for drug-target interaction prediction as well as single-instance ADME tasks, because the system supports multiple task categories including single-instance, multi-instance, and generation tasks.

Related Skills

Explore More Software Engineering Skills→
sunny52julisunny52juli
5

Factor Mining Skill - 重构优化版

Automates creation and optimization of quantitative trading factors using Python libraries.

Community
Advanced
robotlearning123robotlearning123

scikit-learn

Build, evaluate, and refine machine learning models with scikit-learn.

Community
Advanced
Srujan0798Srujan0798

create-viz

Generate Python chart code from datasets using matplotlib, seaborn, or plotly.

Community
Intermediate
ovachieverovachiever
52

pytdc

Provide standardized drug discovery datasets with train/validation/test splits.

Community
Intermediate