Wednesday, September 16, 2009

Human genome variation

I've been reading the latest issue of Genome Research lately, and there's a lot of interesting material in it regarding human genome variation.

First up, Xing et al. look at structural variation created by mobile elements in the human genome. They compared a diploid human genome sequence to the Human Genome Project sequence and found more than 8,000 structural variants. Looking at the variants in the diploid genome caused by mobile elements, they found 706 insertions totaling 305 kb and 140 deletions totaling 126 kb. That's not a lot of DNA (305 kb is about 0.01% of the haploid human genome), but 846 mobile element changes in the human species alone seems like a lot of activity to me. They estimated new Alu elements appear in 1 in 21 births, which is a lot. I guarantee you know someone with a novel Alu element insertion. Probably more than one. And all this variation is likely to have a phenotypic effect, given that Xing et al. found about a third of the indel events occur in genic regions of the genome and three Alu insertions occur in exons. Anyway you slice it, the human genome seems to be changing surprisingly rapidly.

Next, Chun and Fay looked at deleterious mutations in the human genome. There are still creationists out there who argue that all mutations are deleterious, that somehow God made one optimal, perfect genome for Adam (and Eve?), and it's all been downhill from there. (Makes God seem like a kind of incompetent designer doesn't it? Couldn't He have designed something a little more durable?) The actual number of deleterious mutations has not been experimentally verified, and it's not entirely clear how one would go about doing that anyway. In this paper, the authors compared three genomes (Craig Venter's, James Watson's, and a Han Chinese male) to identify amino acid substitutions that are likely to be universally deleterious (i.e., not good in any circumstances). They found only about 800 such substitutions in each person, which is about 15% of the variations present (specifically of the nonsynonymous substitutions). This would seem to be more consistent with a robustly designed genome that is intended to be resilient and tolerant to random changes. This work does not refute the arguments that all mutations are deleterious (for valid technical reasons), but I've always found the all-mutations-are-bad arguments to be ad hoc and resistent to data anyway.

Li et al. found a preponderance of CpG-related substitutions in a survey of variation on Chromosome 21. New structural and sequence variations were found in the first Korean human genome sequence (Ahn et al.). Finally, Shaikh et al. introduced a database of copy number vartiations collected from 2,026 disease-free individuals, which can be found online here.