Back

Natural selections on both amino acid sequences and expression levels are determinants of ohnolog retention

Miao, L.; Li, M.

2021-02-25 evolutionary biology
10.1101/2021.02.24.432806 bioRxiv
Show abstract

The mechanism of ohnolog retention is a subject of concern in evolutionary biology. Natural selections on coding sequences and gene dosages have been proposed to be determinants of ohnolog retention. However, the relationship between the two models is not widely accepted, and the role of regulatory sequences on ohnolog retention has long been neglected. In this study, based on a model of complex traits genetic architecture, we compared the natural selections strength on corresponding sequences between ohnologs and non-ohnologs by comparing complex traits heritability enrichments. We showed that complex traits regulatory sequences heritability enrichments (p = 1.1 x 10-5 in 5 kb flanking regions) and expression-mediated heritability enrichments (p = 2.1 x 10-5) of ohnologs were significantly higher than non-ohnologs. Then, we deduced that regulatory sequences of ohnologs were under substantial natural selection, which was also a determent of ohnolog retention. Meanwhile, we showed that in coding sequences, the complex traits heritability enrichments of ohnologs were significantly higher than of non-ohnologs (p = 9.9 x 10-5), supporting the ohnolog retention model of natural selection on coding sequences. We also showed that complex traits causal gene expression effect sizes of ohnologs were significantly larger than of non-ohnologs (p = 8.8 x 10-6), supporting the ohnolog retention model of natural selection on gene dosages. In conclusion, we provide the first unified framework to show that both amino acid sequences and expression levels of ohnologs are under substantial selection, which may end the long-standing debate on ohnolog retention models.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.