Do I have all the mutations?
September 14, 2021 2:48 PM   Subscribe

I have 6 billion or so DNA base pairs in each cell, and 11 trillion or so cells with DNA in them, and trillions of new mutations every day. What are the chances that all possible single-nucleotide mutations appear at least once in my body?

It looks like about 40% of my nucleotides are C or G, and 60% are A or T. Poking around in a database of mutations, it looks like the relative frequency of mutation types is something like:

A>T: 3.1%
T>A: 3.2%
T>G: 3.5%
A>C: 3.7%
G>T: 4.3%
C>G: 4.5%
G>C: 4.6%
C>A: 4.8%
T>C: 12.2%
A>G: 12.3%
C>T: 21.7%
G>A: 22.0%

I don't quite have the probability chops to put all that together into an estimate of the chances that I have All The Mutations, so I'm hoping that someone in MefiLand does.
posted by clawsoon to Science & Nature (6 answers total) 2 users marked this as a favorite
This is such a great question! But the answer isn't a case of simple probability, as not all mutations in a strand of DNA are equal, and they do not all have the same impact.

Some areas of the genome have higher mutation rates, some areas of a strand of DNA are highly conserved. You are asking about point mutations, when one nucleotide is, through the process of damage and an error in repair or an error in copying during cell division, replaced with another.

Based on where the point mutation occurs in a DNA sequence, it may have a huge/ fatal effect on the cell, or be a silent mutation and the DNA still codes for a functional gene as if nothing changed. One point mutation could turn a gene on or off for example, so if it happens, and that gene is no longer functional, and if it codes for a critical part of the cell machinery, that cell will die. Other areas mutate and that cell sticks around because it doesn't matter.

There is likely no way you could be walking around with a mutation at every nucleotide. If you would like to learn more, I encourage you to read up on coding and noncoding DNA, codons, synonymous and non-synonymous mutations, and point mutations.

This difference in rate of mutations in specific sections of a genome is one of the ways we can use DNA to measure differences between populations or species. Its a really interesting part of genomics!
posted by Maude_the_destroyer at 5:12 PM on September 14 [3 favorites]

Response by poster: I was thinking of asking "virtually all" instead of "all" for all of the reasons you've listed, and perhaps that's the question I should've asked instead. If anyone wants to answer that alternate question, it would be appreciated: Assume that critical mutations are insignificant.
posted by clawsoon at 5:26 PM on September 14

Cells have a lot of machinery designed to prevent mutations, or quickly get rid of mutated DNA (or RNA) that could cause problems. I don’t know enough to give an exact likelihood or even a range, but it’s definitely lower than just calculating number of possible point mutations versus number of cells or number of copies of any given section of DNA.
posted by eviemath at 5:58 PM on September 14

Seems to me the question you're asking is also - or close to - how many mutations does a person accumulate in their lifetime? And then if it's more than, eh, 34 that would put you in "probably" territory for all of them to exist at least once. That may help you.

Of course it's not exactly meaningful because the location of the mutation is way more impactful than what base changed into what other base, but it might be an easier way to get at the question you're asking.
posted by Lady Li at 12:20 AM on September 15

Response by poster: And then if it's more than, eh, 34 that would put you in "probably" territory for all of them to exist at least once.

The minimum number of mutations for the question I'm thinking of would be something like 9 billion. Every one of 3 billion base pairs flips to each of the other 3 possible base pairs. One cell might have one of the 9 billion mutations, and another cell might have another one of the 9 billion mutations.
posted by clawsoon at 10:43 AM on September 15

If you're willing to handwave away all the biology, your question is akin to the Coupon Collector Problem. To make a complete set of 9 billion mutations/coupons/Pokémon, were they equally frequent, you'd have to catch a little over 200 billion on average, which seems doable. If some are an order of magnitude rarer than others, just adjust that figure by a corresponding order of magnitude and you won't be terribly far off.

To save effort, I recommend finding other enthusiasts to trade with.
posted by aws17576 at 8:55 AM on September 16 [2 favorites]

« Older Macbook sound: no output devices   |   What role should I be in UX/UI design? Newer »

You are not logged in, either login or create an account to post comments