This is a shout out to the biologists out there: do you think the concept of dominance and recessive is worthwhile? In other words, does it help in conceptualization more than it hurts? Clearly the idea of recessiveness of deleterious traits helps in comprehending why such alleles exist in the ambient genetic background of a population and can reemerge via inbreeding.1 On the other hand, my own experience is that if you try to move the conversation to additive polygenic traits, which I think are interesting and need to be understood to really "get" population genetics you have to keep batting down the tendency for people to reconceptualize things in a simple dominance-recessive fashion. I think that there must be a distinction between dominance-recessive dichotomies which are only salient at the level of enzymatic concentration, and have almost no phenotypic impact (e.g., the threshold level of enzyme needed for catalysis is almost never lower in the heteryzogous state than in the homyzogous wild type state), and other traits where the character of classification is crucial. In other words, if for example you define "blonde hair" and "non-blonde hair," clearly the latter is "dominant" as a trait. But, if you look at the melanin concentration in individual hairs I am willing to bet that they will reflective additivity and independence. The issue here is human perception and bias of categories. What I'm getting at is that:
a) I am happy to keep the dominance-recessive in cases where the distinction has little fitness implication and the "phenotypic" impact can only be discerned at a biochemical level.
b) But I would like to abandon it in the case of gross and somewhat subjective phenotypes. For instance, curly vs. straight hair. My understanding is that the extent of curliness of your hair is proportional to the shape of your follicle, and it seems upon physical inspection that though children who are biracial from black and white parents have "curly" and "frizzy" hair, its extent is far less than in the black parent. I suspect the follicle shape tends to be equidistant in relation to the parental phenotypes, and the topological definition of "curliness" whould reflect this in the offspring.
1 - The chance of a recessive allele being expressed is the square of its frequency in a random mating population. So, if the population had a recessive allele in a proportion of 0.01, in the next generation 0.0001 would express the trait, or, 1 out of 10,000 individuals. In other words, the overwhelming majority of individuals who carried the allele would not express a deleterious phenotype, so selection is generally very weak on these traits. The inverse occurs with dominant deleterious traits.
- Log in to post comments
I certainly agree that there is much more then dominance vs. recessivceness. What you forgot is intermediate inheritance that can explain many issues without having to assume complex traits.
Indeed, the operationalization of classical genetics has been productive for nearly 100 years and was sufficient for the explanation of the vast majority of human genetic diseases. In addition, without it it would have been impossible to conceptualize complex traits.
One consequence of your suggestion appears problematic to me, though: Everyone is carrying recessive disease causing alleles that may cause problems in the offspring. In addititon you and me may carry alleles that may or may cause health problems when we are aged. So, how will you make a distinction between disease and health? By genetic testing? IMO what counts is the phenotpye. And as long as I am not phenotypically affected I want to live a happy life without thinking of my genetic inventar.
What you forgot is intermediate inheritance that can explain many issues without having to assume complex traits.
you mean additivity on one locus?
Everyone is carrying recessive disease causing alleles that may cause problems in the offspring.
sure. but remember that heterozygosity does not usually imply w = 1, but more like w = 0.9-0.99. across recessive loci this can build up.
So, how will you make a distinction between disease and health? By genetic testing? IMO what counts is the phenotpye. And as long as I am not phenotypically affected
see my previous point. or consider that loss of heterozygosity might be a major issue at t = 65 years even if it isn't for the first 50 years.
dominance and recessiveness are useful because they're simple to understand and many traits follow them to a certain approximation. especially when you're dealing with knockouts, etc, in model organisms, you can make predictions as to the actual molecular functions of a molecule based on the dominance or recessiveness of the trait.
especially when you're dealing with knockouts, etc, in model organisms,
bingo. ok, words like "knockout" or "model organism" suggest me that i am thinking of communicating to lay people. after all, someone that uses the term "knockout" probably knows exactly how to apply the term recessive, and where it is useful. the problem i have had is that once the idea of a recessive and dominant gets nailed into someone's head who doesn't work in biology they start extrapolating and fitting all other genetic systems to this....
The concept of dominance/recessiveness is very badly taught, starting in high school. Most students simply memorize Mendel's 'laws'. If you ask those who've just passed a college genetics courses to define it, they can't, though they can give examples. If you ask them why some alleles are 'dominant' over others, they can't even comprehend the question.
For a while I tried to bypass this problem by starting my genetics teaching with haploid genetics, where phenotype is a direct readout of genotype. But students then became terribly muddled when they tried to make the transition to diploid systems, so I gave that up.
The term knockout is well aqccepted among professional biologists. If you want I may rather use words like targeted mutations, conditional mutants, gene trapping, targeted trapping, ENU, recombinant inbred or recombinant congenic strains. Does this sound more professional? Knockdown and conditional knockdown were't appropriate to describe RNA interference methods established in mice? Mouse geneticists are well aware of things like hypomorhic, balanced and compound mutations, differences of phenotypes in different genetic backgrounds, gene dosis effects, parental inheritance, epigenetics, genetic drift in small caged populations etc.
However, in most instances I was involved in Mendelian genetics were sufficient to describe the underlying traits.
In the rare cases where differences in different genetic backgrounds were observed it was impossible to elucidate the underlying complex traits until now. It is likely that many such differences are due to the fact that any imbred strain contains something like 50 endogenous knockouts that may be necessary for maintain inbreeding. Thus your distinction would not be helpful in these cases. On the other hand we had a case in which different gut flora was responsible for the phenotypic manifestation of a targeted mutations. Still the genetics of the trait was best described as recessive.
However, I must confess that my work is biassed since I mostly work with inbred strains (129SV, C57BL/6, Balb/c) and that your suggestion may be helpful in your studies.
BTW, we used the term "modell organism" in publcations as well as in grant applications and obviously none of the reviwer was ever bothered by this.
The term knockout is well aqccepted among professional biologists.
i wasn't clear. you inferred the inverse of what i was saying.
well I'll chime in...I'm an ecological physiologist and my knowledge of genetics is pretty much limited to what I remember from the only college course I took on the subject (1981! our text--Suzuki?--had really cool little page-flipping animations of meiosis and translation in the margins) and the stuff I teach in intro courses. In such teaching I think you HAVE to start with Mendel's peas, and the straight dominance-recessiveness of his traits are necessary for those first crucial inferences of particulate inheritance, segregation, and independent assortment. I also think it's crucial to then discuss other patterns of expression and then to talk about the mechanisms of dominance and to distinguish between sequence, biochemical, and phenotypic levels of expression, and it's this last part that's always poorly handled in textbooks.
but I have exams to grade...
Sorry, shit happens when you read a post on one day and post a comment later without carefully reading the post or comment you refer to again. Indeed, I even pasted your essential statement:
Sorry again
Martin
no worries :) you misunderstood, but you didn't say anything stupid, so it's all good.