Correlation Analysis of Topological Indices for GERD Drug Models

Mythili V's dataset contains correlation results between degree-based topological indices and model error metrics for Gastroesophageal Reflux Disease (GERD) drug compounds. It evaluates how structural descriptors relate to performance indicators like Mean Squared Error and Standard Error. The data was last updated in April 2026.

Use Cases

• Feature selection for molecular property prediction by identifying which Topological Descriptor correlates most strongly with low MSE.
• Benchmarking graph-based descriptors by analyzing the Correlation Coefficient between different indices and the Property/Target Variable.
• Assessing model uncertainty by examining the relationship between Standard Error values and specific topological indices.
• Validating the relevance of structural features for specific chemical properties by cross-referencing p-values with Correlation Coefficients.

Strengths

• Includes multiple key performance metrics per analysis instance: Correlation Coefficient, MSE, and SE.
• Focuses on a specific, well-defined application domain: Gastroesophageal Reflux Disease (GERD) drug compounds.

Limitations

• Sample size and row count are unknown, preventing assessment of statistical power.
• Geographic and chemical diversity of the underlying molecular graphs is unspecified, potentially limiting generalizability.

Provenance

Source

Mythili V via figshare.

Collection Method

Derived from correlation analysis between molecular graph descriptors and predictive model metrics.

Freshness

Last updated in April 2026.