DDGWizard Integrates Features For Thermostability Prediction
Wang et al. (2025) present DDGWizard, an integrated pipeline published December 1, 2025 that combines 12 computational resources to compute 1,547 features and compiles a feature-enriched dataset of 15,752 ΔΔG measurements for point mutations. They use feature selection to train a predictive model achieving R2=0.61 in cross-validation, outperforming prior ΔΔG predictors, and provide open-source code and data for protein thermostability engineering.
Key Points
- 1Integrates 12 tools to compute 1,547 mutation features across 15,752 ΔΔG measurements
- 2Achieves R2 0.61 in cross-validation and outperforms prior ΔΔG predictors on independent datasets
- 3Enables practitioners to select informative features and predict thermostability impacts for protein engineering
Scoring Rationale
Strong dataset, extensive feature integration, and open-source code provide high utility; limited novelty outside protein-thermostability domain.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
