# What is the statistical relationship between radioactivity and mutation rate?

This question tries to narrow down the scope of that question.

What is the statistical relationship between radioactivity and mutation rate? By how much would the mutation rate be lowered in a idealized world were radioactivity is absent?

According to this wiki page , the average background radiation is 3.01 milli-Sievert per year (including natural and artificial sources). This equals 0.301 rad.

I found a short letter to nature that says the average forward mutation rate in human is 2.6 * 10^-7 per locus per rad = 2.6 mutations per ten million bases. Also it say that this mutation rate is quite uniform among species. The size of the human genome is approximately 3.2 gigabases. So doing a quick maths: (2.6*3.2*100)*0.301 = 250.432 mutations per year per human. This is an approximation because I rounded the size of the human genome to 3.2 gigabases. Also this doesn't mean that we actually acquire this amount of mutations because we have repair mechanisms.

• Interesting stuff + 1 Do you really mean "250 mutation per year per human" or do you mean "250 mutation per reproduction event"? This estimate is huge! A nucleotide typically has a mutation rate of about [$10−7;10−9$]. So random variation of 1 rad (that doesn't seem that huge given the wiki page that you linked) may easily yield to a 10 or even 100 fold higher mutation rate. – Remi.b Apr 4 '15 at 17:26
• You said "this doesn't mean that we actually acquire this amount of mutations because we have repair mechanisms". So they did not take into account repair mechanisms in their study? The letter is 41 years old. I would intuitively think that today's estimate would be lower. – Remi.b Apr 4 '15 at 17:26
• @Remi.b - If you calculate with the mean mutation rate of the interval that you have given (10^-8), that would still give approx 32 mutation events for the entire genome. If we take the upper limit that is 10^-7, we get approximately 320. Maybe using the term mutation event would be better in my answer but I think based on the letter and wiki info that my calculations are correct. But as you say the letter is quite old, I'll check if there's anything new. BTW this would still result in approx (250/(3.2*10^9))*100 = 0.000007826% of the whole genome. – Nandor Poka Apr 4 '15 at 17:50