I was thinking again about the consequences of having more small-effect deleterious mutations than average. I don’t think that they would push hard in a particular direction in phenotype space – I don’t believe they would make you look weird, but by definition they would be bad for you, reduce fitness. I remembered a passage in a book by Steve Stirling, in which our heroine felt as if her brain ‘was moving like a mechanism of jewels and steel precisely formed.’ It strikes me that a person with an extra dollop of this kind of genetic load wouldn’t feel like that. And of course that heroine did have low genetic load, being the product of millennia of selective breeding, not to mention an extra boost from the Invisible Crown.
West Hunter 
This paper makes it look like there’s not much room for mutation load to matter. If two thirds of variation in cognitive ability is additive common variation, and we know there is some nonadditive variation (gene-gene interactions) from twin studies how much might mutation load account for, 5%, 10%?
We’ll see. If you had a deleterious variant that reduced fitness by 1%, on average it would hang around for around 2500-3000 years. There’s high variance in that: most of the alleles in the 1% class would disappear far more rapidly than that, while a few would drift up to fairly high frequencies. I wonder if those relatively high-frequency deleterious alleles might not be fairly well associated with common SNPs.
Well, what does the distribution of fitness burden by frequency look like for deleterious mutations of a given fitness penalty?
It’s proportional to the mutation rate for that class. There is reason to believe that there are more ways to moderately or slightly screw up a protein than to really ruin it, which indicates that mild mutations make up most load in protein-coding sequences. More of the genome is made up of conserved regulatory sequences, but mutations there probably have even milder effects, since few mutations in non-coding sequences cause a serious Mendelian disease.
Amazing timing. Great minds?
Obesity and IQ – JayMan’s Blog
as a father to be in my mid 30s, played by the rules for my entire life (good school, good job, get established, then start a family). wife around same age. After reading this blog for months, I am feeling like I have made a terrible mistake by waiting so long to have kids. Yet, I remain a big fan of this blog. call me a glutton for punishment. Tell me something to feel less bad about the situation.
Mid30s is not that bad – anyhow, what I am talking is mostly the impact of a whole population having old fathers for a hundred generations or more. it adds up. Anyhow, you have to worry more about female infertility at such ages than mutations.
thanks very much for the reply.
to your point about 100 generations – i also wonder what i would want for my own children’s children. play by the “accepted” rules which are not necessarily optimal or do we need to restructure society to encourage teen pregnancy amongst the highly intelligent.
I recently caught up with an old friend of mine, also mid 30s, and he noted the irony that for his entire life we’ve been told to avoid early pregnancy and by the time you are “ready” under the current paradigm it is almost too late. unless you get a bride that is half your age + 7, which is also discouraged by current social mores.
i don’t think current breeding amongst the educated is a pattern which makes any sense, unless we wind up engineering all children in the lab, which barring any radical changes in social organization, seems like it will become the standard for elites as soon as technology allows it.
speed it up, go for twins
I was thinking again about the consequences of having more small-effect deleterious mutations than average. I don’t think that they would push hard in a particular direction in phenotype space – I don’t believe they would make you look weird, but by definition they would be bad for you, reduce fitness.
Can we think of this like a hill in three space, where those with the lowest number of deleterious mutations have the highest Z value and are at the peak of the hill, and those with deleterious mutations are further down the hill?
Those with the average number of mutations would be arrayed around the hill, but about half the way up. Those with fewer mutations would be closer to the top, and those with more deleterious mutations would be closer to the bottom.
Of course, this would really be in n-space, not 3-space.
Amazing timing. More great minds?
Woodley, te Nijenhuis and Murphy (2013) Intelligence, In Press.
Simple reaction times have slowed since Victorian times. Genetic load via dysgenesis?
http://drjamesthompson.blogspot.co.uk/
I suppose I will have to dig into it, but I find that almost impossible to believe. For a couple of reasons.
“There is reason to believe that there are more ways to moderately or slightly screw up a protein than to really ruin it, which indicates that mild mutations make up most load in protein-coding sequences.”
How so? I thought the active sites for most enzymes are highly sensitive to slight perturbations in the amino acid sequence, so much that missense mutations usually destroy the function of the polypeptide altogether. Or at least that’s what I’ve been told.
Perhaps this is true if you factor in regulatory elements and protein subdomains for allosteric regulation. There are also genes coding for structural proteins like collagen, for which I expect to see considerable diversity among human beings.
Hrrrm. Maybe I have some reading to do.
Correction – the missense mutations often do nothing, but whenever they do *something*, the change has got to be substantial!
Shifting the position of a cysteine base by a few angstroms could reduce the catalytic efficiency of an enzyme by several orders of magnitude. Which, in practice, is pretty close to destroying its function entirely?
It’s true that missense mutations in the active site are likely to seriously screw up protein function, but there are orders of magnitude fewer amino acids in the active site than there are in the whole protein.
Take serine proteases, for example. They’ve got a catalytic triad of aspartic acid, histidine, and serine residues that are absolutely essential for function. Change any of them at all, and you completely remove proteolytic activity. But you can change other amino acids without killing the enzyme. Trypsin is normally a selective protease. It specifically targets cleavage of peptide bonds around lysine or arginine residues. But if you mutate Asp189 to Serine, the enzyme still functions, but it loses its specificity and starts cleaving things randomly. Other, even less important residues, may only slightly reduce its efficiency.
In a review article in Nature Genetics, Adam Eyre-Walker and Peter Keightley estimated that ~21% of missense mutations had a fitness impact between 0 and 10-4, 14% between 10-4 and 10-3, 23% between 10-3 and 10-2, 30% between 10-2 and 10-1, 13% between 10-1 and 1.0.
There are lots of ways in which a missense mutation can have intermediate effects. Depends on how important that part of the protein is, how similar the substituted amino acid is, etc.
See here. And remember, these are fitness effects: some proteins are more important than others.
Note also that alanine scanning is a standard technique for figuring out where the important parts of proteins are. Just sequentially replace each amino acid of your protein of interest with alanine. The replacements that break the protein’s activity are important.
Ah, sorry, I was thinking about enzyme function *proper*, not fitness.
And my question was poorly phrased. What I meant to say is that missense mutations that severely disrupt the tertiary structure of the polypeptide are more likely to destroy the catalytic site than not, whether or not you replace a catalytic AA with another.
Now, come to think of it, there are lots of ways missense mutations may allow slight perturbations to the tert structure that may result in gradations of change to the functioning of the polypeptide. There are plenty of examples of this in the literature – not just the one you mentioned about proteases, but others which I ought to have remembered before asking such an ignorant question.
http://drjamesthompson.blogspot.co.uk/ Reaction times are slowing over 120 years. Genetic load?
I agree with misdreavus. Any mutation in a coding gene is more likely to ruin it than not. However, this does not invalidate Greg’s argument. Loss of function genes that would be fatal in homozygotes can be fitness reducing in the heterozygotes. Such genes could hang around for a long time as long as they were rare and purging them meant having two such individuals come together. However, they can be purged in bunches by sexual selection if individuals with high genetic load have trouble finding mates. And of course in hard times even small reductions in fitness can be selected against.
You’re wrong. I study a classic genetic disease, Niemann-Pick type C (N.B., not type A, which is the one that shows up disproportionately in the Ashkenazim). It’s got a prevalence of about 1 in 150,000, and there’s good reason to believe that most of its incidence is caused by de novo mutations. We believe this in part because when we sequence the gene from patients, we find lots of different mutations instead of one or two common alleles, and in part because patients with particularly defective genes die before they turn two – although we’re working on that.
There’s been some good studies trying to correlate the severity of the disease with the location of the mutations, and kids with particularly nasty forms all have either truncation mutations (some codon gets turned into a stop codon, terminating the protein too early) or mutations in two specific subdomains of the protein. Patients with mutations in other regions usually show adult onset problems and have much milder forms of the disease. And interestingly, there are a bunch of regions that we haven’t seen any mutations in.
Since we have no reason to believe that the base mutation rate changes across the gene sequence, that suggests that people with the mutations we don’t find in patients either don’t have problems at all, or more likely, have such mild problems that they don’t notice them or ascribe them to some disease.
Also note: Niemann-Pick C primarily presents as a neurological disorder. In the classical, severe form, it kills off brain cells through excess cholesterol accumulation. In the subclinical forms, who knows? We don’t see those people.
Ok. And you see similar patterns for the myriad dystrophin mutations responsible for muscular dystrophy, a disease which varies considerably in the onset and etiology of symptoms, some of which are not even readily apparent until late childhood. Am I correct?
To the best of my understanding, yes. The difference between (severe) Duchenne dystrophy and (milder) Becker’s dystrophy is largely dependent on how much damage the mutation does to dystrophin. In-frame deletions usually cause Becker’s dystrophy while frame-shift deletions usually cause Duchenne dystrophy. Also, individual IQ in patients correlates with the specific mutations each patient has.
I have wondered if there was some sort of evolutionary tradeoff between muscles and brains over the past hundred thousand years through dystrophin’s dual role. There is some evidence of recent positive selection among proteins that interact with dystrophin, such as DTNBP1 and DTNA.
Any novel environment where higher intelligence can accrue more caloric energy than brute strength alone (see: the invention of the bow) should relax the selection pressure for muscularity. The Neanderthals didn’t fare so well with the brute strength strategy.
Sure: that’s what you might call an inevitable tradeoff, a consequence of the laws of physics. Just as big guys need more food. But because of the way our biochemistry is wired, there can be tradeoffs that exist but are not inevitable consequences of the laws of physics – particularly likely when a gene has two fairly different functions, as they often do.
Yeah, but if old fathering is much of a contributing factor, how come blacks are smarter than the less polygynous subsacharan breeds?
A fair question. I don’t think the old-father effect is the only one in play. However, it’s something we can check out fairly easily.
Might be worth remembering that the Bushmen and Pygmies are outliers to the rest of the human race. Bantu are genetically closer to Europeans or East Asians (in neutral genes) than they are to Bushmen.
east hunter
“I recently caught up with an old friend of mine, also mid 30s, and he noted the irony that for his entire life we’ve been told to avoid early pregnancy and by the time you are “ready” under the current paradigm it is almost too late.”
Part of the reason is education has turned into an accreditation scam. There’s no reason for most people to need a handful of degrees to prove their ability but it makes a lot of money for the gatekeepers.
Indulgences ftw.
I would imagine that this hypothesis, increased mutational load with age would be testable, if not now then in the near future. Take a womens unfertilized eggs or a mans sperm at various ages and see just how much it has degraded at various ages. My question as a curious amatuer is am I right? I would guess that the line for accumulated mutations would not be linear, but would take a sharp dive after age 35, much like fertility for women, but I don’t know. This leads to a second question, are scientists by and large unimaginative chickenshits who rarely ask big questions but instead learn more and more about less and less until they become the world’s leading expert on almost nothing at all.
Women are born with all the eggs they will need in their lifetime. Men do not even begin producing spermatozoa until puberty, and therefore germ line should accrue deleterious mutations with increasing age.
Thanks for your clear answer. Another amatuer question, has evolution found a way of specially protecting the genes in spermatozoa from deleterious mutations for as long as possible.
This is all basic biochemistry.
I am thinking of a number of ways that evolution could select for genes of protective effect, and nearly all of them would either not be worth the cost, or would even diminish the fitness of the bearer far more than the genetic load itself. (For one, you ensure a slower rate of meiosis among spermatogonia. The logic is simple — fewer meiotic divisions = fewer mutations in the haploid cells, but when did producing fewer and _fewer_ sperm ever become a sound evolutionary strategy?)
The problem with genetic load is that 1) while the “bad” alleles might severely impair fitness in the aggregate, they are often close to harmless when examined in isolation, and crossing over will assure that precise combinations of bad alleles will never occur again in another individual (with the exception of MZ twins); in addition, 2) most of these alleles are so rare that selection does a poor job filtering them out of the gene pool, up to the point where they drift to higher and higher frequencies.
You most certainly have a few of these which are unique to you or your extended family members. As for me, I just _know_ I harbor a diverse cornucopia of crappy alleles. Thankfully I won’t reproduce — that’s already been taken care of (ha ha ha!), but my health could be a lot better.
*the germ line
If you’re thinking of chromosomal aberrations like down syndrome that are associated with advanced maternal age, practically all of those result in a severe loss of fitness to the offspring, and therefore should not contribute to genetic load.
Duchenne muscular dystrophy is apparently associated with mental retardation in some patients. (Funny how this was never covered in any of my biochemistry classes. Rote memorization zzzzzz) I already knew that the gene is absurdly long, spanning dozens of exons and multiple promoter regions, but I assumed that the isoform of dystrophin expressed in the brain came from a different, paralogous gene. I was wrong.
Are you implying that genes that influence muscularity AND have pleiotropic effects on cognition (that would be a lot of them!) would not necessarily accumulate loss of function mutations under a slackening of selection pressure — the same way the null allele for ACTN3 has risen to a high frequency among European and Asian populations? (Or, more precisely, that the fitness tradeoffs would teeter on a fine balance, where moderate defects in muscularity might undermine their own fitness value by impairing the brain?)
If so, something like dystrophin might be pretty easy to manage, given the multiplicity of splicing sites and the different structure of isoforms in separate areas of the body.
Duchenne muscular dystrophy drops average IQ by about one standard deviation.
What I was implying was that possibly, increased selection for higher IQ might affect muscle function just because of pleiotropy in the dystrophin complex. It a mutation made you smarter yet weaker, it might fly.
Relaxed selection would, after a long time, leave you with lots of broken versions of ACTN3. There is but one, and you find that null allele in every population on Earth, although frequencies vary. So it’s been around for quite a while. The working versions makes you a better sprinters, the null version gives you more endurance.
You might be right about selection being able to tweak dystrophin for multiple tasks. But that might take a long time: a mutation that generated a net fitness advantage whiling dinging strength might sweep while we were waiting for the optimal mutation. A beneficial mutation carries its neighbors along.
What I was implying was that possibly, increased selection for higher IQ might affect muscle function just because of pleiotropy in the dystrophin complex. It a mutation made you smarter yet weaker, it might fly.
From the paper you published, a lot of the purported high IQ alleles among Ashkenazi Jews are loss of function mutations. (Well, those are just the ones we know about, because only selection can explain the high frequency of the genetic disorders in question.)
I was thinking more along the lines that null alleles would usually undermine the function of an organ, but it seems that’s not always the case. (See: that ugly looking Belgian Blue cow. It seems a lot of these mutations destroy growth inhibitors.) So if the same polypeptide were needed in both the sarcoplasm and the nervous tissue, an impairment in one should result in an impairment of the other.
Or would it?
If there was simply higher mutational load generally speaking rather than relaxed selection for intelligence, we’d expect to see hits to other traits. Does it look that way?
“when did producing fewer and _fewer_ sperm ever become a sound evolutionary strategy?”
When excess sperm is unnecessary? when reduces the chances of fecundation? when it is too expensive to produce sperm?
BTW, viable sperm production is falling everywhere. Sperm banks have to accept counts that a generation ago were considered subfertile. Why?
You have to think in terms of marginal costs. For our species, at least, it is hard to see what benefit it would do.
Or, I should have been more precise.
Excess sperm production indeed incurs fitness costs for plenty of species, but any reduction in spermatogenesis could not possibly arise as a prophylactic measure against genetic load. The whole concept is silly.
Marginal cost, hmm.
Yet excess sperm production must have a significative cost, as our relative the gorilla makes do with less than us and we with less than chimps. There must be some kind of optimization process working. If our women were more faithful, I guess we would need even less sperm than gorillas. We could spend more energy and time exploring the universe.
Would outbreeding have the same effect?
If you take plans from two different clock designs, and build a clock using parts from both, the gears won’t mesh.
What about heterosis (Hybrid vigor).