PredictABEL: Functions for Risk Model Assessment and Simulated Data

A package by Suman Kundu, Yurii S. Aulchenko, and A. Cecile J.W. Janssens provides functions to assess the performance of risk prediction models. It includes measures like c-statistic (AUC), Hosmer-Lemeshow test, net reclassification improvement (NRI), and integrated discrimination improvement (IDI). The package also contains functions for logistic regression analysis and to construct a simulated dataset with genotypes and disease status for evaluating genetic risk models.

Use Cases

• Evaluating model discrimination and calibration based on functions for c-statistic and Hosmer-Lemeshow goodness of fit test.
• Assessing risk reclassification using functions for net reclassification improvement (NRI) and integrated discrimination improvement (IDI).
• Generating simulated genetic and disease status data for method testing based on the included construction function.
• Creating diagnostic plots like ROC curves and calibration plots for model validation reports.

Strengths

• Package includes a wide array of specific statistical measures for model assessment, such as NRI and IDI.
• Functions are specifically written to handle models that include genetic variables.
• Contains utilities for both analysis (e.g., logistic regression) and visualization (e.g., ROC curves).

Limitations

• Row count and dataset scale for the included simulated data are unknown.
• Column-level documentation for any example or simulated data is absent; field semantics must be inferred.
• Last update date is unknown; freshness unverified.

Provenance

Source

paperswithcode

Collection Method

Software package for statistical analysis and simulation.