Jackknife Model Averaging Prediction Methods for Complex Phenotypes with Gene Expression Levels by Integrating...

Area of Science:

• Genetics
• Bioinformatics
• Statistical modeling

Background:

• High dimensional genetic data analysis is crucial for disease risk evaluation.
• Existing prediction methods often overlook inherent group structures in genetic data.
• Incorporating biological pathway information can enhance prediction accuracy.

Purpose of the Study:

• To introduce a novel model-averaging approach, jackknife model averaging prediction (JMAP).
• To enhance high dimensional genetic risk prediction by integrating pathway information.
• To evaluate JMAP's performance against existing methods.

Main Methods:

• JMAP utilizes a novel model-averaging technique for genetic risk prediction.
• It incorporates pathway information into model specification.
• Optimal model weights are selected via a jackknife cross-validation criterion, allowing weights from 0 to 1 without summation constraints.

Main Results:

• JMAP demonstrated superior or competitive performance compared to existing methods like gsslasso in extensive simulations.
• Simulations showed JMAP achieving higher prediction accuracy, with notable gains (e.g., 0.075) over gsslasso under specific settings.
• Real-world application on TCGA cancer datasets confirmed JMAP's comparable or superior performance for continuous phenotypes.

Conclusions:

• JMAP is an effective novel model-averaging approach for high dimensional genetic risk prediction.
• The method successfully incorporates external group structures, such as biological pathways, into model specification.
• JMAP offers improved accuracy and robustness in genetic risk prediction tasks.