Tuesday, January 10, 2012

James Crow colloquium

[[ Embedded version no longer works, but follow link below to view ]]

Excellent colloquium by James Crow (who passed away recently) emphasizing the importance and ubiquity of additive genetic variance. See earlier post -- the paper linked there covers similar material.

@28 min, Nagylaki came to population genetics after doing his PhD under Feynman at Caltech! Here is his tour de force result, mentioned by Crow in the talk. Interested physicists, see also here.


Justin Loe said...

Neher at UCSB also did this paper that I came across yesterday. Interesting modeling of the effects of epistatic interactions on recombination and recombination maps. 

highly_adequate said...

Crow makes the point that the missing heritability might well be due to a multitude of genes of very small effect -- too small, generally, to be detected by the means available at the time of Crow's lecture (late 2010).

What I didn't see in the Or paper, in my very cursory look at it, is any reason to believe that such an account isn't viable. I got the impression that the argument was mostly model based: they could construct a model in which epistasis acted to create "phantom" heritability, and so just went to town with it. But does it have anything to do with the real world? Why might we regard the model of a large number of genes of small effect as inadequate?

On the other hand, I guess I can see how disease risk might be different. If a disease typically consists in the disruption of normal pathways of relatively small length, then perhaps there are relatively few ways those can be so disrupted, and so one might expect fewer genes to be playing a role.

Dima Klenchin said...

Nagylaki came to population genetics after doing his PhD under Feynman at Caltech! Here is his tour de force result

Quite a contrast to the math autodidact Motoo Kimura who founded Neutral Theory.

Richard Seiter said...

Inspiring to see someone still making contributions in his 90's.

NYT obituary: http://www.nytimes.com/2012/01/11/science/james-f-crow-population-genetics-pioneer-dies-at-95.html

steve hsu said...

Crow was right on top of things until the very end. He notes correctly in the lecture that

1. Population genetics has until now produced a lot of theory but very few testable predictions.


2. This will all change with the advent of cheap sequencing. He points his younger colleagues in the right direction!

Richard Seiter said...

To retain that much intellectual power and flexibility at that age (and a sense of humor ;-).  One can only hope...

Thanks for the link.  There were quite a few nuggets in the talk.

Justin Loe said...

To the extent that epistatic interatiotions exist there is an evolutionary tradeoff:
" the impact of epistasis and pleiotropy on adaptive evolution...The highest fitness is attained in landscapes with an intermediate amount of ruggedness that balance the higher fitness potential of interacting genes with their concomitant decreased evolvability. "
ref: http://www.ncbi.nlm.nih.gov/pubmed/21697174

Justin Loe said...

Clarifying the point for myself: beyond and upper bound, the amount of epistasis is less adaptive: (from the paper above)
"The attained fitness is maximal at K ¼ 3 to 5, from which we infer that an intermediate amount of epistasis and pleiotropy is most conducive to adaptation"

using the NK model: "NK model of genetic interactions [31–33] consists of circular, binary sequences encoding the alleles at N loci, where each locus contributes to the fitness of the haplotype via an interaction with K other loci."

Blog Archive