Many plant scientists will create experimental populations of siblings, so called “F1 populations”. They use these to find locations on the genome important for different traits, like yield and flavor. Because of the law of Mendelian segregation (which you probably learned in high school), researchers typically assume that the distribution of offspring genotypes (what their DNA looks like) is something known. They thus often test the observed offspring against this theoretical expectation. At a locus on the genome where there is a large deviation from what’s expected, researchers take this as evidence of “segregation distortion” and sometimes discard that locus. They thus use these tests for segregation distortion as part of a quality control pipeline.

However, polyploids (organisms with more than two copies of their genome) exhibit more complicated segregation patterns than diploids (like us humans). They exhibit patterns of “double reduction” and “preferential pairing” (two processes in polyploid meiosis), which can alter the expected distribution of offspring genotypes. Polyploid genotypes are also often more uncertain than diploids. Both of these issues make common tests detect too many loci in segregation distortion.

In our new preprint (Gerard et al., 2024), we develop a new model for meiosis in tetraploids (organisms with four copies of their genome), and use this new model to account for double reduction and preferential pairing in tests for segregation distortion. We also optionally account for genotype uncertainty. Both of our modifications result in tests that don’t detect segregation distortion too often.

Check out our preprint:

Gerard, D., Thakkar, M., Ferrão, L. F. V. (2024). Tests for segregation distortion in tetraploid F1 populations. bioRxiv.