boron-nmr-predict

Predict 11B NMR chemical shifts from SMILES using local CPU inference.

52|11|Updated Mar 10, 2026

Search Tags:#cheminformatics #huggingface #smiles #cpu-inference #boron-nmr #nmr-prediction #molecule-image

Ready to run

npx skills add https://github.com/InternScience/ChemClaw --skill boron-nmr-predict

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: boron-nmr-predict
Source: https://github.com/InternScience/ChemClaw/tree/main/skills/boron-nmr-predict
Command: npx skills add https://github.com/InternScience/ChemClaw --skill boron-nmr-predict

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires huggingface_hub, rdkit-pypi, Pillow, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Boron-containing molecules often require experimental data or expensive quantum calculations to assign 11B NMR shifts. This skill provides an offline, CPU-based workflow to predict boron-11 chemical shifts directly from input SMILES and present a clear, labeled visualization.

Core Features & Use Cases

Local CPU inference with a 5-fold ensemble for robust boron-11 NMR predictions.
Accepts SMILES inputs, downloads model weights from Hugging Face on first run, and outputs per-boron ppm along with a labeled molecule image.
Includes environment setup, solvent handling, and optional ML features to enhance prediction accuracy.

Quick Start

Create or activate a CPU-based environment, then run predict with a boron-containing SMILES to obtain per-atom ppm predictions and a labeled image.

Frequently Asked Questions about boron-nmr-predict

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

FAQPage Schema

How do I predict 11B NMR chemical shifts from a SMILES string on CPU?▼

To predict 11B NMR chemical shifts, provide a boron-containing molecule SMILES to run local CPU inference, which outputs per-atom ppm values and a labeled molecule image. Model weights download automatically from Hugging Face on first use.

Can I predict boron-11 NMR shifts offline without quantum calculations?▼

Yes, you can predict boron-11 NMR shifts offline using a local 5-fold ensemble model. It avoids expensive quantum calculations by running entirely on CPU inference after the initial Hugging Face model weights download.

Does this 11B NMR prediction tool require a GPU or conda environment?▼

No GPU is required; 11B NMR prediction runs entirely on CPU inference. The workflow supports conda-based environment setup to manage dependencies like rdkit-pypi and Pillow for generating labeled molecule images.

What is the best way to visualize predicted NMR ppm values on a molecule?▼

The best way to visualize predicted NMR ppm values is by generating a labeled molecule image. The workflow maps chemical shift predictions directly to boron atoms using RDKit and outputs the image file path for review.

How does the boron NMR prediction model handle solvent effects?▼

The boron NMR prediction model includes solvent handling and optional ML features to enhance prediction accuracy. This approach refines the chemical shift estimates from the 5-fold ensemble based on the molecular environment.