Comparative Genomics Approaches Accurately Predict Deleterious Variants in Plants

Area of Science:

• Genomics
• Evolutionary Biology
• Plant Science

Background:

• Genome resequencing advances drive interest in predicting functional consequences of genetic variants.
• Variants at conserved sites are more likely to impact fitness and phenotype.
• Current prediction methods are primarily validated on human disease mutations.

Purpose of the Study:

• To assess the accuracy of predicting deleterious genetic variants in plants.
• To evaluate the performance of comparative genomic approaches in a nonhuman species.
• To understand the implications for plant evolution, domestication, and crop improvement.

Main Methods:

• Generated a curated database of 2,910 Arabidopsis thaliana mutants with known phenotypes.
• Evaluated seven different comparative genomic approaches for predicting deleterious variants.
• Compared the performance of these approaches against human benchmarks.

Main Results:

• All evaluated approaches performed well in predicting deleterious variants in Arabidopsis thaliana.
• The relative ranking of these approaches differed significantly from their performance in human studies.
• Deleterious mutations can be reliably predicted in Arabidopsis thaliana.

Conclusions:

• Deleterious mutation prediction is reliable in Arabidopsis thaliana and likely other plant species.
• The predictive performance of various approaches is species-specific and does not directly translate from humans.
• This work has implications for evolutionary studies and crop trait improvement.