R1b1 aka P25 and reversion


John McEwan

3rd June 2006 (revised 23rd Oct 2006)



P25 is a SNP that defines the haplogroup R1b1 in the ISOGG 2006 Y chromosome tree and is widely used by FTDNA. There are a number of issues about this SNP that are continually being raised on the Genealogy DNA list server and this page is an attempt to summarize them. Please remember that P25+ notation is the mutant or positive form and P25- is the ancestral variant.


Genographic project calls individuals M343+ when they have actually been genotyped for P25

The Genographic project calls individuals R1b (M343+) when in fact they have been genotyped for P25 by FTDNA who do the genotyping for this project.


This was described in an email by Bennet Greenspan on 10th Oct 2005 to David Reynolds subsequently posted on the Genealogy DNA listserver.


On Mon, 10 Oct 2005 12:21:24 -0700, Bennett Greenspan <bcg@familytreedna.com> wrote:

Hi David
We actually test P-25 for all FTDNA clients and all NGS clients. The
NGS folk list those folks as M343+ but since we test P25 (and because
P25 is below M343) we test the same for both and NGS refers to this as
M343. Hope this helps.

Does this matter? In many cases no, but in perhaps 2% of cases it provides an incorrect answer (see below). If comprehensive SNP testing is undertaken this can often be identified, but if selective SNP testing is used this is not the case. Summary, if you are using Genographic reported SNPs be alert to this problem.


P25 is an unreliable SNP and should be treated with caution

It is commonly assumed that when a SNP mutates, for most genealogical purposes it is what is called a “unique event polymorphism” or UEP. What this means is that it happens only once in one individual and it does not revert. This assumption is nearly but not quite correct, because each individual has about a 1 in 50,000,000 chance of mutating at that SNP. Given the number of people in the world some must have mutated. However, in the past the world population was much smaller and this assumption is reasonably reliable. In the Y chromosome tree some SNPs have reverted, but most have not. When they have reverted they have been typically widely separated in space and time and are easily recognizable.


P25 is a SNP but is located in an unusual portion of the Y chromosome which has been duplicated 3 times. A diagram of the P25 region is shown in Fig. 1 below from Adams et al (2006) who investigated reversion this SNP. Such duplicated regions can be subject to a process called “gene conversion” or “recombinatorial loss of heterozygosity” or RecLOH for short. This process essentially swaps the duplicated regions and its signature is loss of any variability that was present between the regions. The process is explained in Wikipedia. The consequence is that while this is still a rare process it can be much more frequent that mutation of a SNP.



Adams et al. tested 1000 males from the Iberian peninsula and found 10 M173(xSRY1532b) P25- individuals (1%) and 22 out of 421 in the British sample (5.2%). Of these 31/32 were M269+ and one was also SRY2627+ (in the paper given as M167+). What does this actually mean? If you look at the ISOGG tree closely and follow where the SNPs are in the tree it means almost all and perhaps all of the 32 P25- identified had “reverted” at some time in the past to the ancestral form. In practice it means that using P25 as a marker alone will result in 1-5% of individuals being misclassified as to there position within the tree. Any result from P25 should be viewed with caution.


Recently, additional information has become available from publicly available genealogical testing. The R1b SNP table on this site  has 5 P25- out of 245 P25 tests that are otherwise consistent with being R1b (2.0%) and this estimate is in a mid range estimate of the above 2 populations. The combined results from both sources also show that there have been at least TWO reversions: one in SRY2627+, and a separate one in S21+.  What is still not clear is what number, if any, have occurred in R1b1c(xSRY2627,S21). Previously a Y STR network diagram based on 19 STRs in Adams et al (2006) provided only equivocal evidence of multiple founding events and the variability observed of the reversions was similar to R1b as a whole (inferring that the change was very old or there were multiple events). Given that we detected P25- in S21+ the significantly higher level in the "Great Britain" sample from Adams et al. may be because of the percentage of S21+ in this population.


It has been estimated that gene conversion events occur much more frequently than SNPs, but perhaps at a 10X lower rate than STR mutations (1 in 500 events). If so the number of reversion events may be modest and some lineages could be of reasonable size. The new EA SNPs may be able to resolve this issue and identify at least some of the founding events.


The R1b SNP list also has a result Denney RS4EF and Davis 4TWBD, which have not yet been tested for S21. Such test would confirm a possible additional reversion events if they were S21-. I would certainly recommend that any individual who is M269+ P25- also tests for at least S21, S28 and SRY2627. The outcome would be a diagnostic clan SNP profile especially when paired with STR markers.



Adams, S.M., King, T.E., Bosch, E., Jobling, M.A. 2006. The case of the unreliable SNP: recurrent back-mutation of Y-chromosomal marker P25 through gene conversion. Forensic Science International 159:14-20