MetaQTL

MetaQTL integrates results from gene mapping experiments to derive consensus Quantitative Trait Loci (QTL) and marker positions and to merge genetic maps for analysis of the genetic basis of complex traits.

Key Features:

• Modular Library: Implemented as a modular set of programs written entirely in Java.
• Data Integration: Integrates molecular markers, QTL, and candidate genes across multiple studies.
• Statistical Framework: Establishes consensus models for marker and QTL positions across the entire genome using a comprehensive statistical process.
• Consensus Map Creation: Merges distinct genetic maps into a single optimal consensus map using weighted least squares and supports investigation of recombination rate heterogeneity between studies.
• Clustering Approach: Applies Gaussian mixture models to determine the number of QTL underlying observed distributions.
• Performance and Validation: Validated by simulations showing that standard model choice criteria from mixture model literature perform effectively and that clustering can reduce confidence interval length for QTL location given sufficient observed QTL across studies.

Scientific Applications:

• QTL meta-analysis: Combine QTL results from multiple studies to identify consensus QTL positions.
• Consensus genetic map construction: Create an optimal merged genetic map from disparate marker maps using weighted least squares.
• Recombination rate heterogeneity analysis: Investigate differences in recombination rates between studies via consensus map comparisons.
• QTL number determination and localization: Use Gaussian mixture models to infer the number of underlying QTL and refine their locations.
• Integration of candidate genes: Relate candidate genes to consensus QTL and marker positions across studies.

Methodology:

Implemented in Java; integrates molecular markers, QTL, and candidate genes; establishes consensus models for marker and QTL positions genome-wide; merges maps using weighted least squares; performs clustering with Gaussian mixture models; validated by simulations using standard model choice criteria.