A recent estimate of human mutation rate:
I have been saying here now for more than a year (and more like 10 years over all) that if people marry who are not sufficiently related they will suffer a loss of fertility that will accumulate over generations as similar mismatches occur.  The rationale behind saying this is that there are things called mutations, random little changes in the DNA.  Since the DNA of any animal must be fine tuned, that fine tuning is degraded, which is manifested as a fall in fitness, particularly in fertility.

Depending on that logic, I built the computer program I have described before to look at the fertility of a virtual population with virtual genes and virtual mutations.  The program is quite successful, explaining a number of things that people have noticed and been pretty well baffled by and some things people have not noticed. 

But the question does arise, “How many mutations?”  This has left me feeling rather uneasy.  Many years ago, when I was a student at Harvard Medical School, we had a lecture in which somebody had calculated what he thought must be the mutation rate for humans, how many new mutations each human had.  That was decades before we could analyze a whole genome, so he had to rely on indirect means, such as looking for the effects of mutations and comparing mutation rates of different kinds of animals.

In those days it took an average of about 6 months for a healthy woman to become pregnant.  He calculated that an average person should have about 4 new mutations.  In order for the genome to be at equilibrium, those 4 mutations would have to be removed from the gene pool.  He reckoned they were removed 2 at a time, but even so, there would be more babies lost carrying off the mutations than the number of babies who would develop.  This, he proposed, was why it took more than a month for a woman to conceive.

I had not heard anything that seriously challenged his numbers even though of course there was an enormous amount of new information piling up over the years.  So in building my computer model I sought to keep the number of mutations per individual very low.  However in my most recent efforts, in which I increased the maximum number of individuals in a population and drastically increased the number of potential offspring (before allowing for those lost to mutations) I found that I had to postulate a fair number of mutations, a lot more than 4.

I now see (Y-rated, NATURE vol. 461 no. 7260, September 3, 2009 page 15) that an article has been published ((Curr. Biol. doi:10.1016/j.cub.2009.07.032(2009)) that estimates the mutation rate.  They looked at the Y chromosome of 2 Chinese men separated by 13 generations and found 4 mutations, suggesting a mutation rate for the whole genome of 100 to 200 mutations per person per generation.

I am sure that number will be refined by looking at more individuals and by finding out whether every base pair is as likely to undergo mutation as any other base pair.  Nonetheless, I think it is safe to say that the number is “a lot more than 4.” 

This is encouraging in a couple ways.  For one thing, it is nice to see that the computer program, as I tweaked it, forced me into making an assumption that then seemed implausible but now seems quite reasonable. 

For another, assuming the genome to be mostly at around equilibrium, 100 to 200 mutations is simply too much for humans to survive, even if they are carried off 2 at a time.  If a woman is fertile for 25 years and ovulates 13 times a year, that would make 325 maximum ovulations.  If 50 offspring must be lost for one to survive, that would give an absolute maximum birth rate of 6 to 7 offspring.  But if there were 6 pregnancies lasting 9 months, that would mean 54 missing ovulations or 251 total ovulations and a maximum of 5 offspring per woman average.  There are populations that have far exceeded that.  Besides it would mean it would take years not months to get pregnant. 

In other words, the impact of a single mutation must be less than ½ of 1 offspring.  This is exactly what the model assumes.  So even though my work continues to be ignored despite my most strident efforts, I think there is hope that somebody else will get here.  Once they get a better handle on that mutation rate and begin to calculate the impact of one mutation, it should not take them long to realize that fine tuning exists and that outbreeding will result in problems that accumulate over genations.

However I suspect that there is actually another level of complexity in real life and that this other level is in fact what the computer program is modeling.  I’ll tell you about it some day when you are older. 

There have been 2,675 visitors so far.

Home page.