I read online that they have found around 10 million single nucleotide polymorphisms (SNPs) across individuals in the human genome [1]. However, this number includes all SNPs found in all people who have been sampled and does not tell us much about SNPs in any single person.

I was wondering instead how many SNPs there are in a single person on average. I had trouble finding trustworthy evidence about this question.


3 Answers 3


I was wondering how many SNPs there are in a single person on average

A SNP is a polymorphism in the population, it is not a thing a haplotype can carry. Each individual has a given variant for any given of the SNPs (except for cases of sequence deletion).

It is possible though to say how many SNP a diploid individual in between its two haplotypes but I doubt this is where you interest was. Ewens smapling formula gives expectation and distribution under an infinite alleles model, panmictic population and absence of selection.

Other statistics that make sense include the average number of deleterious mutations pre individual or a few other things or how many new mutations a couple transmit to their offspring,... But asking how many SNPs an individual carry on average makes little sense.

Note as @Chris said, by stating 10k SNPs in the human genome, you seem to under-estimate the number of SNPs.


In the comments, you further

Would it make more sense to consider a rough minimum and maximum?

The answer is no, it would not make more sense. Consider the definition of SNP by wikipedia

A single nucleotide polymorphism, often abbreviated to SNP (pronounced snip; plural snips), is a variation in a single nucleotide that occurs at a specific position in the genome, where each variation is present to some appreciable degree within a population [My emphasis]

The term population is key here. The concept of SNPs (or of polymorphism to be more general) makes only sense if you can make comparisons. A single individual does not have any polymorphism (to the exception of the one between his two sets of chromosomes).

If we step back and think of different species it might become more intuitive. Think of the number of pairwise differences between two species. It does not make any sense to ask "how many pairwise differences is there between a blue whale?" You need to say "how many pairwise differences is there between a blue whale and a wolf". Similarly, you cannot say "How many SNPs is there 'between a single individual'" (to the exception again of the consideration of the differences between the two sets of chromosomes), you need to consider a group of individuals.

  • $\begingroup$ "But asking how many SNPs an individual carry on average makes little sense." It does make sense. Heterozygosity in individuals is a useful proxy for genetic diversity in the population. $\endgroup$ Commented Sep 6, 2016 at 15:48
  • $\begingroup$ We can sure compare the two haplotypes (as I said in the second paragraph). This is not heterozygosity (nor observed heterozygosity) but just an average number of pairwise differences between haplotypes within each individuals. I just doubt this is what the OP was after but maybe I am wrong in the OP's intentions. $\endgroup$
    – Remi.b
    Commented Sep 6, 2016 at 15:54
  • $\begingroup$ My field of expertise is Computer Science. I have started a bioinformatics project which input would be a vcf document of SNPs (of a human individual). I'm interested in the average for practical and optimisation reasons, rather then biologic scientific reasons. But as you stated, it makes no to consider the average. Would it make more sense to consider a rough minimum and maximum? $\endgroup$
    – BARJ
    Commented Sep 12, 2016 at 2:25
  • 1
    $\begingroup$ @BARJ Please see edit. Hope that helps. $\endgroup$
    – Remi.b
    Commented Sep 12, 2016 at 3:01

There was only one source I was able to find, It stated that each genome contains 1 SNP every 1000bp. (Take a look here). However I have no idea if this source is reliable


Your number is too low, the 1000 Genomes project lists roughly 15 Mio SNPs (see reference 1). To find out how much SNPs humans have on average (this depends of course on populations, where this can differ), you often find that on average 1 in 1000 nucleotides is altered if you compare two human genomes (or the other way round: They are 99.9% identical).

The problem with this number is that it is quite old and can be traced back to a publication from 1991 (see reference 2), which compared 75.000 sequences nucleotides from 49 loci. Since this was before the complete sequence of the human genome was known, I would treat this number with some caution.

The newest number I found on this is from publication of the human genome. They estimate the number a little bit lower, at about 1 in 1300 bases (see reference 3).

What could be done with the data from the 1000 Genomes Project is to download the data from the individual samples, map it against the reference genome and see, how many SNPs per individual occur. I am not sure if nobody did that or if I simply overlooked that.

Based on this data there are somewhere between 2,3 and 3 million SNPs (1:1000 or 1:1300 respectively) present in each human genome. According to the 1000 Genomes project (reference 1), the rate of newly occurring SNPs is $1\times 10^{-8}$, meaning that there are between 20 and 30 new SNPs occuring per generation.


  1. A map of human genome variation from population-scale sequencing
  2. Low nucleotide diversity in man.
  3. Initial sequencing and analysis of the human genome.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .