LMM-Lasso

LMM-Lasso combines linear mixed models and Lasso (Least Absolute Shrinkage and Selection Operator) regression to perform multi-locus genetic mapping and multi-marker phenotype prediction while controlling for confounding effects such as population structure.

Key Features:

• Multi-Locus Mapping: Maps multiple genetic loci simultaneously to capture effects of numerous interacting genetic factors.
• Correction for Confounding Effects: Controls for confounders such as population structure to reduce spurious associations and false positives.
• Parameter-Free Approach: Operates without tuning parameters while maintaining statistical power across genome-wide datasets.
• Phenotype Prediction: Facilitates multi-marker-based phenotype prediction from genotype data.
• Scalability: Scales to large genomic and genome-wide association study (GWAS) datasets.

Scientific Applications:

• Arabidopsis thaliana GWAS: Applied in GWAS of Arabidopsis thaliana to obtain significantly more accurate phenotype predictions for 91% of the phenotypes studied.
• Mouse linkage mapping: Used in mouse linkage mapping to dissect phenotypic variability into components attributable to individual single nucleotide polymorphisms (SNPs) and population structure.

Methodology:

Integrates linear mixed models with Lasso regression to simultaneously model multi-locus effects and control for confounding variables, enabling identification of likely causal variants and multi-marker phenotype prediction.