The frequency of the genes (alleles) for each marker can provide useful information when interpreting results of the Y-chromosome DNA test. Thanks to the Blair DNA project, we have information on the frequency of each allele from over 4000 tests for the first 12 markers, and more than 2500 tests for the second 13 markers.
    In the table below I have extracted the frequency for the alleles found in our group of DNA participants. For each DYS marker, there may be one or several alleles (values in the table below). For example, at DYS 393 we have values of 12, 13, and 15; at DYS 454 the only value found among our group is 11. Below each row of DYS values are the frequencies of each of those values from the Blair DNA analysis.
    If all of the marker values for a person have a frequency of 20% or greater, there will be DNA matches to people with other surnames, where no relationship is likely within the last 500 years. Look at our family group #3. Most of their values have a frequency of 50% or higher. Their rarest marker is a value of 14 at DYS437 which has a frequency of 20%. If we consider only the first 12 markers, they match 200 people with surnames other than WEST. They have a very common DNA profile or haplotype. It is possible, even likely, that they share a common ancestor with some of those other surnames, but the ancestor pre-dates our genealogy time frame. That ancestor lived before surnames were adopted.
      W04 is at the other extreme in allele frequency. He has two very rare alleles, 19 at DYS 385b and 33 at DYS 449. Each of those alleles has a frequency of 1%. Either would make his haplotype rare, but having both makes his haplotype very, very rare. He has zero matches to anyone of any surname.
    These frequencies can aid us in evaluating DNA matches to people with other surnames. If you match someone with another surname on 11 of the first 12 markers, or 23/25 or 24/25 on the 25 marker test, and want to examine the possibility of whether you are related, compare your DNA haplotypes. If you and the other person share a rare marker, for example you both have a value of 10 for DYS 385a, and you differ at a marker such as DYS 439 with values of 11 (32%) and 12 (48%), it may be worthwhile exchanging information with the other person. The first step would be to determine from your paper genealogy whether you and the other person had ancestors who lived in the same locallity at the same time. If you have a close match to someone with another surname, but the difference is at a marker where one of you has a rare allele, then the probability of a relationship is lower.

   If there is a match to a male with a different surname on they Y-chromosome DNA test, there are several possible explanations. Perhaps we are conditioned by our times to jump to the conclusion that there has been a "non-paternal" event, that is, the father of record was not the biological father. While that is certainly a possibility, and is known in many cases, other events can lead to the same result. The first was mentioned above, two people with different surnames could have a common ancestor before surnames were adopted. Another common event in our early history was adoption and changing of surnames without official records. Take the example of a husband dying young and leaving a widow with small children. The widow could marry again, and the children take the surname of the new husband. The care of orphaned children was often directed by local courts, but there may be examples of these children changing surnames without official records. Children abandoned in infancy, might have no record of their parents, and the surname could be assigned or assumed. We are also aware of men who changed their surname and started new lives, keeping their former identity secret. The DNA of their descendents could match the DNA of a descendent of their brother or uncle, who passed down a different surname. There is also the possibility of genetic "convergence".  This results from the fact that our DNA "mutates" or changes over time. If mutations did not occur, we would all have the same DNA and we would all be genetically identical - what a boring world that would be! If we could test the DNA of two unrelated men who lived in the year 1000, their DNA might differ on 6 of the DYS markers. During the 40 or more generations from that time to today, mutations would have occurred in the descendents of these two men. Mutations are random, we cannot predict their occurance, only draw general conclusions based on thousands of observations. It is possible that male descendents of these two men would have matching DNA today, resulting from the random nature of the mutation process.

    We may also find cases where a paper genealogy indicates that two men have a common ancestor, but the DNA test tells us they are not related. Some people will believe that the DNA test is faulty and their paper trail is correct. This could be checked by having another DNA test performed by a different lab. There are now 3 or 4 labs doing DNA testing for genealogy, and several people have had their DNA tested by all of these labs. So far, all of the labs have returned the same result for individuals who have tested at multiple labs.
     If the DNA results are valid, the paper genealogy should be re-examined. Most of us know of cases where the wrong person has been placed in a family based on circumstantial evidence. If this has occured, then the descendent of that person would not match a descendent of another branch of that family on the DNA test. If the paper genealogy appears faultless, with good documentation at each level, then a break in the surname line should be investigated. This break could be the result of events mentioned previously; a "non-paternity" event (the father of record is not the biological father); an undocumented adoption resulting in a surname change; a male who changed his surname and kept his former identity secret etc.
    Occasionally the paper trail will find what we assume are unwed mothers, females for whom there is no apparent marriage record, but with children who have the surname of the mother. We also find females who are listed as a head of household that includes children with different surnames. We may assume that these are her birth children, but we cannot find a marriage record to a male with the surname of the children. In these examples, a DNA test may offer the only method of identifying the father of these children. If the researcher can find clues in the record as to the possible father of the children, it may be possible to find descendents of that line. One of those descendents could be asked to have his DNA tested, to compare with the DNA of the researcher who is trying to find the male ancestor.

---