Using the rough estimate of 7,000 centiMorgans in a human... Is there an estimate that breaks down this total by chromosome (including the sex chromosomes)? E.g., in the human karyotype, do chromosome pair 1a and 1b have n1 cM, 2a and 2b have n2 cM, etc. so that n1 + n2 +...+ n22 + sex chromosomes = 7,000?

If you understand what I'm getting at and can re-phrase my question to clarify it, please do so.


The short answer is "yes". Here is an early example: https://science.sciencemag.org/content/265/5181/2049.abstract

I am sure that with some searching around you can find the most recent version. I was able to find somewhat more recent maps here. The data has to be in the UCSC genome browser somewhere. Most likely you should find your 7000 cM estimate source and figure out which map that estimate comes from if you like that source.

These data structures are called "genetic maps" or "linkage maps". Each chromosome has positions that are defined by cM, and you could use that for your purpose I believe by just taking the largest cM value marker for each chromosome.

(Note that the two copies of a chromosome do not have independent genetic maps- after all, the map is defined by recombination between the two homologs. e.g. for chromosome 1 there is usually just one map, and therefore one maximum cM value- the map is an average across the human chromosomes sampled.)

Edit The question seems to be related to a genealogical heuristic for identifying relatedness in cM values. I'm still not 100% sure how they measure this exactly.

The source shares expected values for relatedness here: https://secureservercdn.net/

| improve this answer | |
  • $\begingroup$ The short answer suits my purposes well, as I am trying to get a rudimentary grasp on the mathematical and biological factors that apply to genetic genealogy. I understand that parent/child relationships are a special situation, and with all other familial relationships "damned statistics," recombination, etc. are strong factors. For some reason, self-reported sibling relationships in a 25,000+ person study by Blaine T. Bettinger are averaging significantly lower (about 2600 cM shared ) than the expected 3500 cM for a 50% DNA match. Can you direct me to a source with a layman's answer? $\endgroup$ – Joseph Filip Jan 8 at 0:53
  • $\begingroup$ @JosephFilip it would help to see the actual source to evaluate this. I am not sure what it means to share cM between sibs as it's a measure of physical linkage on chromosomes rather than genetic relatedness. $\endgroup$ – Maximilian Press Jan 8 at 6:04
  • $\begingroup$ Edit: it appears to be related to this: thegeneticgenealogist.com/2018/12/17/…. I'm not a genealogist per se, I guess this measure kind of makes sense in that context, though it's unfamiliar to me. The layman's answer might be that blog post. $\endgroup$ – Maximilian Press Jan 8 at 6:09
  • $\begingroup$ @JosephFilip Note: sibs aren't 50% matches, but rather a bit less than. Parent-child are 50% because of Mendelian inheritance (one chr copy from each parent), but sibs are comparing the outcomes of two such random draws (one for each sib). In other words the values you note 2600 cM are expected for sibs (see the chart linked in answer/blog post). see here for more details: hoylab.cornell.edu/relatedness.html $\endgroup$ – Maximilian Press Jan 8 at 6:18
  • $\begingroup$ To a hammer, everything looks like a nail. To a 1970s math major, everything looks like an equation. I did my own analysis of the grandparent/grandhild inheritance probabilities, but I can't figure out how to post it. I used the assumption, known to be wrong, that chromosomes behave as identical immutable coins being tossed. My calculations for averages and "normal ranges" match with "summary info" on genealogy and DNA test websites. They give a figure but don't say how it was derived. The hoylab site says that sibs SHOULD BE 50% matches. btw I do know that cMs are a rate not a distance. $\endgroup$ – Joseph Filip Jan 9 at 1:51

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.