After doing some statistical analysis of the raw math involved in DNA contributions, I would like to understand the scale of the biological factors. I know that 99% of the human genome is identical in all human beings, so genetic recombination and gene conversion should be irrelevant in those areas. I know that recombination occurs at different rates depending on the chromosomes and loci. But what is the average order of magnitude? That is, if I know that a chromosome mainly came from Grandma, how much of it is likely to be from Grandpa? .1%, 1%, 10% ? And, does the age of the male affect the rate of recombination to a measurable degree (in spermatozoa only, as body cells would not affect inheritance).

  • $\begingroup$ What do you mean by "the scale of the biological factors"? And what do you mean by "the average order of magnitude"? As you just said, differences are tiny between any two randomly drawn humans (in DNA sequence), so if this is your 'metric' there can't be any differences in 'orders of magnitude'. If you only ask about the fraction of the particular DNA (in chromosomes) that was copied into your own genome from your ancestors, then it is roughly half from your parents, a quarter from your grandparents, and so on. $\endgroup$ Commented Oct 5, 2019 at 21:27
  • $\begingroup$ DNA tests are done only on the variable segments: It would be pointless to say "you're a 99% match to this other sample, it must be your identical twin." Using the "average" of the actual distribution is misleading. Only 32.24% of people inherit between 23.91% to 26.09% from a grandparent while 90.69% inherit somewhere in a range of 17.39% to 32.61%. $\endgroup$ Commented Oct 6, 2019 at 1:54
  • $\begingroup$ Agree, so, what is that you are asking then? $\endgroup$ Commented Oct 6, 2019 at 1:56
  • $\begingroup$ DNA testing is in its infancy, and different labs use different protocols for their tests and different models to report results. 2 samples from ME will not show a 100% match to MYSELF. The zygote that turned into my mother, say, had 23 intact chromosomes from each of her parents. When my zygote was formed, recombination may have occurred in the specific ovum that made me. Take chromosome 1- the main body of which came from her father. Is it likely to be 99.9% grandpa with .1% recombined from grandma, or more in the 90% grandpa 10% grandma range? $\endgroup$ Commented Oct 6, 2019 at 2:25

1 Answer 1


So, at core (it seems) you are interested in the rate of recombination events and the magnitud of their occurrences. This would be the only source I can think of that would give you information about the 'composition' of one chromosome (other than brute-force sequencing).

As you seem to know, recombination is an important factor that 'shuffles' some specific areas of the chromosomal DNA during meiosis (process needed to produce gametes). There is a technique developed by T. H. Morgan almost a century ago, that measures the rate of this 'recombination' events by comparing how two (or/and more) particular loci co-ocurr across generations (when recombination occurs, it is more likely to break co-occurrences of distant loci, as opposed to close loci). Thus, we can use this distance measure as a measure of recombination. In fact, in honor of Morgan, the units used to measure recombination rates are called 'centiMorgans', and abbreviated cM. A cM is the probability for a particular loci (it can be a gene or a smaller sequence) to co-occurr with another loci (to be 'linked' to). Thus, is a relational measure: every particular sequence is related to another, and by transitivity (if A and B are close, and B and C are also close, then A and C will probably be close -this is a generalization-). When done in chromosomal or genomic scale, we can build linkage maps.

In modern times, there's quite more resolution, from whole genome studies at gene/chromosome resolution, and even higher resolutions (remember that a 'loci' can be any unit of your interest in the DNA).

Being all that said, recombination rates, and thus the probability of inheriting certain DNA features from parents/ancestors based on recombination, vary quite a lot. There are many 'hot-spots' and other regions seem to be 'protected' from recombination (for instance, regulatory DNA). According to the work I cite here, the probabilities of recombination can vary within about an order of magnitude (the most likely recombination occurring at 10 times higher probability than the less likely), but this varies a lot.

I hope this helps a bit. I am still not sure what do you mean by "biological factors", but for sure recombination is one of the important factors influencing the identity of each chromosome across generations (others are, of course, point and chromosomal mutations).

  • $\begingroup$ I reviewed my understanding of centiMorgans. It appears to be the measurement of a rate, but the units used are a distance. To me, this is like estimating travel time by considering only the distance, with no regard as to terrain, weather or traffic. As you said, there are known hot-spots in the genome. With a cM being a .01 probability, and rounding to about 7,000 cMs in a human, that would be an average of 70 instances of genetic recombination in a single generation, and that's a non-directional switch, as opposed to conversion. With variables you point out, only general ranges matter. $\endgroup$ Commented Oct 6, 2019 at 4:50
  • $\begingroup$ It gives a rate because the probability of 'linkage' between any loci can only be in the [0,1] range. So, even though it is thought of as a 'distance', in reality (both as conceived by Morgan, and in real 'physical' terms) it reflects the frequency of recombination events. So, to me, it is a bit like the inverse of your metaphor: we get traveling times, yet we like to talk about it as a 'distance traveled'; the weather, traffic, etc, are implicit to the traveling times observed (the linkage). $\endgroup$ Commented Oct 6, 2019 at 4:58

You must log in to answer this question.

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