QTL mapping for seed morphology using the instance segmentation neural network in Lactuca spp.

Area of Science:

• Plant genetics
• Agricultural science
• Bioinformatics

Background:

• Wild lettuce (Lactuca sp.) domestication likely began for oilseed traits.
• Seed morphology is vital for oilseed quality, but its genetic basis is poorly understood.
• Manual phenotyping of small lettuce seeds is challenging for genetic studies.

Purpose of the Study:

• To develop and apply an automated image analysis pipeline for quantifying lettuce seed morphology.
• To identify quantitative trait loci (QTLs) associated with seed shape traits in lettuce.
• To investigate the genetic basis of seed shape variation during lettuce domestication.

Main Methods:

• Developed an instance segmentation-based pipeline for automated seed morphology quantification.
• Performed quantitative trait locus (QTL) mapping using ddRAD-seq data from a cross between oilseed and wild lettuce.
• Analyzed 7 seed traits: area, width, length, length-to-width ratio, eccentricity, perimeter length, and circularity.

Main Results:

• Identified 11 QTLs linked to 7 seed morphology traits.
• Discovered three significant QTLs (qLWR-3.1, qECC-3.1, qCIR-3.1) on linkage group 3 (LG3) associated with seed shape.
• These QTLs are located in a region known for domestication-related genes, suggesting selection for oilseed traits.

Conclusions:

• Domestication of lettuce for oilseed involved selection for specific seed shapes, driven by genes in the identified genomic region.
• The study provides genetic evidence supporting the oilseed type as a driver of lettuce domestication.
• Automated image-based phenotyping is effective for genetic dissection of small morphological traits in plants.