We are all half dead!

This will be the first post in a new series on this blog, it will be in English and deal with papers, accompanying a course on Quantitative Genetics for Master Students. I am only voluntarily participating in this courses and try to keep up with paper discussion/essays this way...
...I will try to make the discussion both accessible for the interested reader and deep enough to meet the criteria of a good essay for the course.

Back to the topic. What do I mean with the title of this post?
With "we" I mean all animals and with "half dead" the fact that the average animal carries (heterozygote) more than one recessive allele that would be lethal if homozygote. The lethal allele has no influence on fitness of heterozygotes (half dead means fully alive here;-)) and reduces homozyygote fintness to 0.

You don`t need fancy technology to figure this out and the methods used for the study of McCune et al. are nearly as old as the field of quantitative genetics itself (the reference describing the method is in fact from 1927 and not accessible online). The experimentator mates simply siblings resulting from a cross of wild-caught animals and records the zygotes or embryos with developmental distortions leading to death. Sofar this seems facile, the only difficulty in interpretation of the reults is easy to resolve: If similar phenotyps are observed in different crosses the phenotypically healthy siblings from both crosses are outbred with each other. When all the ofspring of this controll is healthy two different recessive lethal allels were found.
This method has a single severe downside: In animals that have a reduced rate of survival as embryos or zygotes due to chance or environmental influences the experiments are not possibele. Therefore Xenopus laevis was the only vertebrate for which data on R (the number of recessive lethals per individual) was investigated by this method before. Extensive data is in contrast available on R for Drosophila.

McCune et al. found, that in both teleost fish species (Lucania goodei and Danio rerio) R is of comparable size to R in Drosophila.
The title "A Low Genomic Number of Recessive Lethals in Natural Populations of Bleufin Killifish and Zebrafish" is already part of their interpretation. They suggest that vertebrates have a higher number of genes and therefore R is smaller in relation to the number of sites that can cause lethal phenotypes.

McCune et al. infer the number of genes in their fishes from the number of genes in humans. They postulate that, because of the high synteny between vertebrates this number would be approximately the same. Unfortunately the estimation for the number of human genes was 35,000 back in 2002, the real number based on newer estimates is not higher than 25,000. The ratio of Drosophila/vertebrate genes comes down from 2.5 to 1.79 considering this.
McCune et al. propose the smaller size of vertebrate populations as a reason for the lower R (in relation to the number of genes) in vertebrates compared to invertebrates. This sounds intuitively right, because smaller populations result in higher inbreeding. For this reason selection against recessive lethal alleles would be more effective in the smaller vertebrate populations.

Nevertheless I have other doubts regarding the plausibility of the assumptions that R must be set in relation to the "exome-size". There would be no need for correcting with the number of genes, if all animals had a set of genes comparable in size, essential for their development. As R is the same in all animals the whole discussion (and the title) of the paper would make no sense in this context.

P.S. I am not aware whether vertebrates and invertebrates have this comparably large set of essential genes. It seems not to be known yet...
...or I should search harder.


Amy R. McCune, Rebecca C. Fuller, Allisan A. Aquilina, Robert M. Dawley, James M. Fadool, David Houle, Joseph Travis, and Alexey S. Kondrashov (2002) A Low Genomic Number of Recessive Lethals in Natural Populations of Bluefin Killifish and Zebrafish. Science 296 (5577), 2398.
[DOI: 10.1126/science.1071757]