How many alleles per gene the human species have today?

[jaumzaum]

Hello. Yesterday I was making some calculations, and I started to wonder the average number of alleles per gene the human specie have today. The best question would be how many mutations per base or how many alleles in an average 1000 base gene, as genes vary in size.

What made me wonder this?
Sometime ago I saw a 1200 base carcinogenic gene that had 31 mutations documented. For me this is actually different from 31 alleles as someone could have more than one mutation, resulting in 2^31 = 2 billion possible alleles. And we also have alleles that were not documented. As the chance of a specific gene suffer mutation from all the genes we have is low, I would say that more than 1 mutation has a very low chance. And 50-100 alleles would be a good guess. Let's take 60. That would give 1/20 mutations per base in our genome (e.g. considering all the 7 billion people living today, we would expect to find 1 mutation in each 20 bases).

But we know the mutation rate in germinative cells is 2.5 10^-8 per base per generation. Considering the human specie began 1 million years ago, and a 20 year generation period, we would expect 1/800 mutations per base in our genome.

So, where is the mistake or the wrong assumption?
I thought about 2:
The first one is that we considered the mutation rates for germinative cells, and the gene I considered was a somatic carcinogenic gene that could and possibly had a higher mutation rate than germinative cells. That leads me to the previous question, what is the right number of mutations per base we have today considering all the human population (and the mutations we are born with) , or the average number of allele per gene?
The second is that in the second calculation we considered a model in which all the humans came from 1 and only 1 ancestor. Something like 1 pre-human gave birth to all the human that ever existed. This is because we did not consider that the first ancestors already had different genes one another (i.e. mutations). If we considered this, it is not right to calculate the number of mutations since human specie started as there were mutations even before.

Can someone help me to solve this paradox?

 

[jim mcnamara]

I think you are overdoing your analysis. A decent definition of allele:
"one of two or more alternative forms of a gene that arise by mutation and are found at the same place on a chromosome."

How many loci?

There are an estimated 20,000-25,000 human protein-coding genes. The estimate of the number of human genes has been repeatedly revised down from initial predictions of 100,000 or more as genome sequence quality and gene finding methods have improved, and could continue to drop further.

- from ScienceDaily

There are 324 million known variants. (See link below)
Average calculation $$324000000 / 25000 = 12960$$ - IMO, this is a meaningless number, however it answers your question. A more useful answer: lots of alleles.

Notice that having an allele does not necessarily change phenotypes we see, that is not mentioned in the definition. Alleles are all alternatives, not all death sentences.

"Differences between individuals, even closely related individuals, are the key to techniques such as genetic fingerprinting. As of 2017, there are a total of 324 million known variants from sequenced human genomes."

from: https://en.m.wikipedia.org/wiki/Human_genetic_variationAlso:
Consider only existing alleles, many of the precursors (those that came before) may not have existed in any organism for thousands of years. Or possibly never.

 

[BillTre]

I am not going to give you a number, but I will try to provide some relevant information.

Species is both the singular and plural of the word.

Several words can have different meaning depending on who uses them in which situation. These are related terms:
Mutation:

• a mutation can be an event that causes a change in a gene
• a mutation can be the result of a change in a gene (I used to mutagenize things, so I am inclinded toward this kind of meaning. An mutagenizer can make mutations of already mutagenized genes with various methods for various purposes, for example a revertant, mutating back to "normal".)
• you seem to be using mutation to describe every little difference in a gene.

Allele:

• Created alleles: I take this word to mean a different version of the gene in question that is found at some time, often after mutagenesis (in which each new version is a new allele (even a reverant which would mutate a mutant back to a "normal" allele).
• Found alleles: In my view, a found allele (one that is noticed in a survey of a population) could have several sequence differences from"normal".

Version or Variant:

• I would use version as the result of several mutations on a single gene, or single mutations.
• As well as in the cases of found alleles, which could (or might not) have several differences from "normal".

As @jim mcnamara pointed out not all gene variants are bad. For example silent mutations are nucleotide sequence changes that do not change the the amino acid sequence of the protein produced and therefore rarely have any noticeable effect.
In addition, there are mutations that do not effect gene sequence. These could control sequences (like promoters and enhancers), but could also effect the distances between control these elements (making them too far away or too close for them to work "normally".

The second is that in the second calculation we considered a model in which all the humans came from 1 and only 1 ancestor. Something like 1 pre-human gave birth to all the human that ever existed. This is because we did not consider that the first ancestors already had different genes one another (i.e. mutations). If we considered this, it is not right to calculate the number of mutations since human specie started as there were mutations even before.

You are right to question this. A species does not normally evolve from a single ancestor, or even two. It is usually a population of many individuals (could be thousands or more), although in unusual case it could be one (such as parthenogenic species). This starting population could bring in several different alleles of any gene. If they are already there in the species's founding population, would they still be considered mutations?

Also as @jim mcnamara pointed out, molecular methods can now be used to actually count the number of sequence differences in various parts of a gene or the whole genome. This would probably be the modern approach which could give you the best numbers for your question.

 

[Ygggdrasil]

Here are two useful reference for you to read regarding mutation rates that touches upon some of the questions you bring up:
http://book.bionumbers.org/how-genetically-similar-are-two-random-people/http://book.bionumbers.org/what-is-the-mutation-rate-during-genome-replication/
Here is a relevant excerpt from the second article:

Given the existence of these various mechanisms of genome rearrangement, it is interesting to consider the extent to which the space of possible genomic mutations is explored. A recurring class of estimates in various contexts, such as the famed Levinthal paradox, center on how well biological systems “explore” the space of all possible outcomes. In many of these examples (protein folding, space of possible genomes, etc.), the astronomical numbers of possible outcomes are simply staggering. As a result, it is easy to wonder how thoroughly the space of possible mutations is “searched” within the human population. We explore how such an estimate might go in Figure 5. Given that there are about 7 billion people on earth, with on the order of ≈10 mutations per generation, we estimate that the current human occupants of the planet explore roughly 7×10⁹x10 ≈ 10¹¹ new mutations during the turnover from one generation to the next. This means that if we focus our attention on any single site within the 3 billion base pair human genome, dozens of humans harbor a mutation in that particular site. As a result, the space of single base pair mutations is fully explored amongst the entire population of humans on earth. On the other hand, if we consider a specific two base pair mutation we find that by random mutation it would require on the order of 10⁷ generations of the human population to achieve it by chance!

Why are estimates of the number of alleles of genes much lower than we would expect from these calculations?
1) Selection. Many mutations in genes are lethal. Embryos containing such mutations are not viable and are miscarried. These lethal mutations help account for the fact that 30-50% of all fertilized eggs are eventually miscarried.

2) Common vs rare variants. Most databases of alleles and genetic variants consist of ~thousands of individuals and therefore are missing much of the extant genetic variation in the human population. We have a pretty good catalog of common variants that resulted from mutations many generations ago and are therefore present in mahy millions of extant humans. However, we have not sequenced many of the rare variants (potentially present in only one individual on Earth) that exist in the human population.