Talk:Gene-environment interaction

From Wikipedia, the free encyclopedia

	This article is part of WikiProject Genetics, an attempt to build a comprehensive and detailed guide to genetics on Wikipedia. If you would like to participate, you can edit this page, or visit the project page to join the project and/or contribute to the discussion.
???	This article has not yet received a rating on the quality scale.
???	This article has not yet received an importance rating.

==Psychology bias==

This entry is way too psychology-biased at the moment. I'm fairly certain that gene-environment interactions were talked about in biology several decades earlier than the article currently says of psychology (1990s; biology: 1950s? I'm thinking Waddington). - Samsara 16:03, 13 January 2006 (UTC)

I would have guessed JBS Haldane... Pete.Hurd 06:01, 28 January 2006 (UTC)

Contents 1 Merge? 2 Example to add 3 Pre- and post-natal 4 Heritability 5 removed chart

[edit] Merge?

While we're at it, could this page not be merged/redirected to heritability or nature versus nurture? it's really short, and I'm not sure where it's intended to go that would cover ground not properly covered on those pages. Pete.Hurd 06:01, 28 January 2006 (UTC)

[edit] Example to add

Francis et al 2003, Nature Neuroscience 6:445-446.

[edit] Pre- and post-natal

I wonder if there isn't a way to make a few remarks on the distinction of pre-natal and post-natal environments. Either one clearly falls in the category "environment", but I think many readers will tend to forget about pre-natal environment as a entering into the interactional quality of development. Maybe some reference to the Francis, et al. paper that you sent me, Pete.Hurd (Epigentetic sources of behaviral differences in mice), which was interesting in considering both environments along with gene strains. Lulu of the Lotus-Eaters 21:24, 31 January 2006 (UTC)

Yeah, I'll put some thought into some text, I don't have time to make a nifty graphical description of the Francis paper, but I think it would be super-useful to triple underline the idea that "born that way" does not equal "genetics" due to the importance of uterine environment effects. Maybe some mention of how MZ twins that share a single amniotic sac are more similar than MZ twins that each had their own... I'll see if I can find that reference too. Pete.Hurd 23:01, 31 January 2006 (UTC)

[edit] Heritability

The concept of heritability really needs working in, somewhere in the first paragraph. Pete.Hurd 05:13, 3 February 2006 (UTC)

[edit] removed chart

I removed the chart, even though it is a really nice picture, because it does no necessarily give a correct description of GxE interaction. An interaction would occur if the same environmental manipulation has differential effects on individuals (plants in this case) with different genotypes. If the environmental effect is the same regardless of genotype, we have si,ple additivity of genotype and environment. As an example, take two genotypes G1 and G2 and two environments E1 and E2. Additivity would occur if, for example, we have the following case:

G1E1 = 10

G1E2 = 15

G2E1 = 20

G2E2 = 25

(E2 increases the studied character with the same amount in each genotype). An interaction occurs in this example:

G1E1 = 10

G1E2 = 11

G2E1 = 20

G2E2 = 30

(E2 increases the character in each genotype, but much more so in G2 than in G1). Or, in the case of Tryon's rats:

G1E1 = 10

G1E2 = 20

G2E1 = 20

G2E2 = 20

(G1 and G2 differ in E1 but not in E2). --Crusio 16:59, 21 September 2007 (UTC)

Huh? Obviously the chart doesn't illustrate every possible norm of reaction. But the effect of environmental change is still, of course, differential across genotype. The point of the chart isn't that "all gene/environment interaction is exactly this", but rather "interaction can be in a manner that contradicts stereotypes about the meaning of heritability". That's a point worth adding... I'll improve the caption. LotLE×talk 17:42, 21 September 2007 (UTC)

The chart still is not correct. My point is that the chart could simply indicate a situation where plants in the right hand environment grow, say, 10 cm less tall than in the left hand environment, whatever their genotype. That is not genotype x environment interaction, but additivity of genotype and environment (my first numerical example above). GxE is when different genotypes react differently after a particular environmental change. Such as Tryon's rats, where the Bright line was the same, whether they were reared in an enriched or an impoverished environment, but where the Dull line changed its behavior dramatically when the environment changed. That is interaction. Under both additivity and interaction can you have a situation where heritability (within an environment) is 100% and different environments produce different phenotypes. With additivity this would always be the case. With interaction, you could have examples where heritability in one environment is 100% and 0 in another. Hope this clarifies, I realize this is getting rather technical.

I leave the chart for the moment, I don't want to start an edit war :-) but I think it should be taken out as it does not give a correct illustration of what GxE interaction is. --Crusio 19:24, 21 September 2007 (UTC)

So, if the chart were changed to identify different genotypes within each treatment, and then show that the effect of the environmental variation was different across the environmental change, then we'd be ahead. (?). First glance I took at the figure, I assumed we were supposed to think that the leftmost plant was the same genotype, etc... Pete.Hurd 19:33, 21 September 2007 (UTC)

I did not create the chart, so am not certain what it's creator intended. But my reading was that each pot illustrated seven different genotypes, arranged in the same order in the left and right pots. The response to the changed soil is not identical across genotypes: for example, the tallest plant on the left does not match the position/genotype of the tallest one on the right. The point in the graphic caption is that within each pot, all the variation is genetic (since the environmental conditions are identical within a pot. Between pots an environmental effect changes expression of the group (and only environmental, since the same genotypic range occurs in each pot). LotLE×talk 19:52, 21 September 2007 (UTC)

In order to demonstrate that the variation between plants within treatments is entirely genetic there will have to be replicates showing that different individuals of the same genotype have identical heights (under the same environmental conditions). Pete.Hurd 19:56, 21 September 2007 (UTC)

It depends what we're trying to show, doesn't it? If genotype A in environment 1 has variation in expression... well, I guess that still means <100% heritability, right? Do you think expanding the box caption can clarify it well enough, Pete.Hurd?

Btw, I modified the image to make explicit my reading about the seven same genotypes in each pot. I created a PNG version because the format is more appropriate, but it also means the JPEG is not modified. LotLE×talk 20:22, 21 September 2007 (UTC)

Given the subject matter of the article, I think we might as well make the figure illustrate a nice GxE interaction. The figure makes no distinction between noise and genetics, I like its potential, but I think the figure itself will have to be changed before it makes a worthwhile pedagogical payoff. Pete.Hurd 20:51, 21 September 2007 (UTC)

I agree with Pete. In fact, I think that the current figure is too complicated. The point that within a certain environment all differences between individuals can be 100% heritable, but that despite that a change in environment can still modify the phenotype of these individuals is a valid one, but does not belong here. It would be better at its place in the article on heritability. There it should be modified also, to show that heritabilities are population and environment specific. To get back to Tryon's rats, under normal circumstances the trait "learning ability" is highly heritable, but under enriched conditions the heritability is essentially zero. --Crusio 14:15, 22 September 2007 (UTC)

If it's the shared opinion of Crusio and Pete.Hurd that this isn't the right chart for this article, I bow to that consensus. I realize it's not a "fully general" illustration of the concept. Maybe if Pete or Crusio has an idea of what a better illustration would be, you could write to the creator of the first chart to see if he could do it. I like the style of the plants of different heights, in any case: it just feels friendlier for this than a completely abstract bar chart or line chart. LotLE×talk 18:23, 22 September 2007 (UTC)

I agree that it's a much friendlier chart, and with tweaks it will work great to illustrate the topic. I'll look more into it's creator, and drop some suggestions and see if we can coax further labour from them. Pete.Hurd 19:45, 22 September 2007 (UTC)