Machine Learning Performance Metrics for GERD Drug Compound Analysis

Performance metrics for machine learning models used in QSAR/QSPR analysis of Gastroesophageal Reflux Disease (GERD) drug compounds. The dataset, created by Mythili V, was last updated in April 2026. Each row corresponds to a model, detailing its algorithm, target property, and validation scores.

Use Cases

• Benchmarking model algorithms like SVR and Linear Regression against metrics such as R² (test) and RMSE for a given Target Property.
• Analyzing the impact of different Feature Set / Descriptor Type choices on validation metrics like MAE and Q².
• Evaluating the consistency of model performance using the reported Cross-validation Strategy (LOOCV) and Training Metrics R² (train).

Strengths

• Includes multiple validation metrics per model (R², RMSE, MAE, Q²) for thorough evaluation.
• Specifies the cross-validation strategy (LOOCV) used for model assessment.
• Released under a permissive CC-BY-4.0 license for reuse.

Limitations

• Unknown row count prevents assessment of statistical power for comparative analysis.
• Lacks raw compound structures or descriptor values, limiting reproducibility of the modeling.
• The dataset's scope is limited to GERD drug compounds, which may restrict generalizability.

Provenance

Source

figshare, author Mythili V.

Collection Method

Presumably compiled from the author's own QSAR/QSPR modeling experiments on GERD compounds.

Freshness

Last updated in April 2026.