README.md

metadata

license: cc-by-4.0

Nanobody Thermal Stability Dataset

Dataset Overview

This dataset helps predict how stable nanobody sequences are at different temperatures. Thermal stability is important for nanobody engineering and applications, affecting how well they work in different environments.

The dataset includes two types of stability measurements:

• Melting temperature (Tm): The temperature at which nanobodies start to unfold
• Sequence stability: Stability scores based on sequence properties

Data Collection

This dataset comes from experimental measurements of various nanobody sequences. The data is collected from published scientific literature and laboratory measurements, then clustering based split.

Dataset Structure

The dataset is split into training, validation, and test sets:

• train.csv, val.csv, test.csv

File Format

Each CSV file contains these columns:

• seq: Nanobody amino acid sequence
• label: Thermal stability value (melting temperature or stability score)

Uses and Limitations

Uses

• Develop machine learning models to predict nanobody thermal stability
• Help design more stable nanobodies
• Provide reference data for nanobody research

Limitations

• Limited dataset size may not represent all nanobody families
• Experimental conditions may affect measurements
• Models should account for data distribution characteristics

Evaluation Metrics

Model performance is evaluated using:

• Spearman correlation
• R²
• Root Mean Squared Error (RMSE)
• Mean Absolute Error (MAE)