Analyzing Medicago spp. seed morphology using GWAS and machine learning.

Area of Science:

• Plant genetics
• Agronomy
• Bioinformatics

Background:

• Alfalfa (Medicago sativa) is a crucial forage crop with diverse seed characteristics.
• Understanding the genetic underpinnings of seed morphology is key for crop improvement.

Purpose of the Study:

• To investigate the morphological diversity of alfalfa seeds.
• To identify genetic markers associated with seed morphology traits in Medicago spp.

Main Methods:

• Phenotypic screening of 318 Medicago accessions for seed traits.
• Genome-wide association study (GWAS) using SNPs against the Medicago truncatula genome.
• Machine learning models for validating associations and identifying complex trait markers.

Main Results:

• Significant phenotypic diversity and correlations among seed morphology traits were observed.
• GWAS identified candidate genes (CPR1, MON1, PPR, Wun1) near associated markers.
• Marker S7_33375673, upstream of Wun1, strongly predicted red color intensity and brightness.

Conclusions:

• The study elucidates the genetic basis of seed morphology in Medicago spp.
• Identified genes and markers provide valuable resources for breeding improved alfalfa varieties.
• Correlations among traits suggest pleiotropic effects influencing seed development.