I have heard that humans share 99.9% of their DNA with other humans. I have also heard that a child shares 50% of their DNA with their parents. How do I resolve this apparent contradiction? It has been really bothering me.

  • 1
    $\begingroup$ There was another recent question about this that gave an answer that is equivalent to the answer for your question, asking about Neanderthal DNA in Homo sapiens sapiens, see here: biology.stackexchange.com/questions/61147/… The short version is: they are talking about slightly different things. $\endgroup$
    – Bryan Krause
    Jun 29 '17 at 0:21
  • 1
    $\begingroup$ See here: biology.stackexchange.com/questions/41974/… as well $\endgroup$
    – Bryan Krause
    Jun 29 '17 at 0:24
  • 1
    $\begingroup$ And here we go with that 99.9% again: biology.stackexchange.com/a/61078/24284 $\endgroup$
    – user24284
    Jun 29 '17 at 6:15
  • 2
    $\begingroup$ Why was the question down voted? I think it is a common misunderstanding. We have already seen such misunderstanding several time on Biology.SE but it is the first time that the origin of the misunderstanding was so clearly phrased. I think it is a good question. $\endgroup$
    – Remi.b
    Jun 29 '17 at 22:06

It will be clear with a simple analogy.

You are 50% related to any one of your parent

Let's say you don't have any biology books. You have two friends, Alice and Bob. They each give you a copy of the book Campbell Biology. You now have two Campbell Biology. You have received 50% of your Campbell biology books from Alice and 50% from Bob.

Similarly, you inherit 50% of your DNA from your mother and 50% from your father. You are related at 50% to any one of them.

Two randomly sampled individual are 99.9% identical

Now consider the list of all the copies of Campbell Biology in the world. As there exist different editions, all Campbell Biology won't exactly be the same. Let's say you randomly sample two Campbell Biology from around the world and you align them letter by letter. What is the expected fraction of the letters that will be identical? If, for example you find the two sentences

Selection is a fitness variance associated to a genetic variance among individuals in a population.


Zelection is a fitnezz varianze azzociated to a genetic varianze among individualz in a populazion.

There are exactly 9 mismatches out of 99 characters, that is a 90% similitude.

Similarly, if you randomly sample two humans, align their DNA (a DNA sequence look like ATTTCGCTGTCGAATCGATCGGTA), you'll find that the fraction of mismatch is lower than 0.1%. Therefore, we all share 99.9% of our DNA.

Of course, alining DNA sequences (or normal sentences) is not quite that easy as some sequences (or sentence) can have more nucleotides (or letters) than others but I won't go into the details here.

How do these two measures relate?

Let's say that instead of giving you a book, Alice actually produced a copy of its book and gave it to you. The book you have received from Alice (that is 50% of all your Campbell Biology books) is 100% identical to Alice's book in term of mismatch.

Similarly, the 50% of your DNA that you inherit from your father (or mother) is 100% identical to the copy of the genome found in your mother.

Note however that

  1. Your mother probably made between 10 and 100 mutations when copying her DNA so that the DNA you received from your mother is not exactly 100% identical to your mother's DNA

    • Similarly Alice could have miscopied her Campbell Biology book that she passes to you
  2. Your mother actually recombined her two haplotypes

    • Similarly Alice actually had two different editions of Campbell Biology and she mixed them up a little bit before copying the resulting book!
  • 1
    $\begingroup$ Good answer. +1 But could you add a link between the two? E.g. that the 50% you have from your parents are 100% identical to the respective parent in terms of 'shared mismatches' (disregarding the 60-100 de novo mutations) and that because of this the expected identity between parents and offspring is way higher than between randomly sampled individuals. $\endgroup$ Jun 29 '17 at 10:05
  • $\begingroup$ @AlexDeLarge Very good point, thank you. I tried to add that. Feel free to rephrase it if you think you can make it easier to understand. $\endgroup$
    – Remi.b
    Jun 29 '17 at 10:21

The 99% and 50% refer to different senses of relatedness.

Humans share 99% of genes with other humans: If you were to compare one human to another human with respect to every single one of their genes, you would find a 99% similarity.

A child shares 50% of his or her genes with his or her parents: This 50% refers to a relatedness relative to the population rate of relatedness. It actually refers to a gene within that 1% that is different (a polymorphism) between humans. This is an approximation used in order to calculate what degree of prosociality an individual could be selected to impart to another individual, where the gene for prosociality would have some frequency between 0 and 1 (that is, there would be some difference between individuals at the locus where the gene is located). Hamilton's (1964) rule is used for this calculation. It is c < b*r, where c is the cost of the actor; b is the benefit to the recipient of the act; and r is the relatedness between the actor and recipient. The r used in this rule between parent and offspring is 0.5. The interpretation is that if a gene is responsible for a cost (c) to the holder of the gene (a parent) and a benefit to some other individual (an offspring), the relatedness (r) between the holder and the other individual must be greater than c/b because relatedness gives the likelihood that a copy of the gene responsible for the cost and benefit is also located in the benefiting individual and so the likelihood that the offspring shares a prosocial gene carried by a parent is half the difference between 1 and the background rate. If the gene is at 70% frequency in the population and one parent has the gene, then the offspring's likelihood of having the gene is 85% (assuming the other parent has it with 70% likelihood).

  • 1
    $\begingroup$ "Humans share 99% of genes with other humans." This is just wrong. Healthy humans share every single gene, i.e. 100%, with every other healthy human. The 99%+ refer to sequence identity, not to gene identity. -1 $\endgroup$ Jun 29 '17 at 9:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.