CVRMS

CVRMS selects optimal subsets of genetic markers from rank-based datasets such as genome-wide association study (GWAS) results or marker effect analyses to enable genome-wide prediction of phenotypes using cross-validation and ridge regression in R.

Key Features:

• Marker Subset Extraction: Identifies subsets of genetic markers from rank-based datasets (GWAS or marker effect analyses) that are most predictive of phenotypic traits.
• Ridge Regression Optimization: Uses ridge regression to address multicollinearity among predictors and improve prediction accuracy of selected marker subsets.
• Cross-Validation Framework: Implements cross-validation to evaluate predictive performance and generalizability of marker subsets.
• Heritability Range Handling: Applies to datasets with heritability values ranging from zero to one and to data from humans, animals, and plants.
• Large-Scale Biomarker Selection: Selects from hundreds to thousands of biomarkers for genome-wide prediction.

Scientific Applications:

• Genome-wide Prediction: Facilitates prediction of phenotypes from genome-wide marker data using selected marker subsets.
• Complex Trait Analysis: Supports investigation of genetic contributors to complex traits across varying heritability.
• Cross-Species Genetic Studies: Applies to human, animal, and plant datasets for genetic research, breeding, and medical genetics applications.

Methodology:

Processes rank-based datasets from GWAS or marker effect analyses, applies cross-validation, and fits ridge regression models to select predictive marker subsets, outputting subsets of hundreds to thousands of biomarkers.