Testing for ySNPs is the new frontier in genetic genealogy for direct
paternal ancestry. Its utility is improving rapidly and it may (possibly)
replace STR testing. In the meantime, it has become a valuable supplement to
This page tries to provide a guide for the average genetic genealogist.
What are ySNPs?
A ySNP is a SNP on the Y chromosome and a SNP (pronounced "snip") is a
single nucleotide polymorphism at a particular position on the chromosome in
which one base (A, C, T, or G) is replaced by another. A test for a ySNP is
either negative (the usual case) or positive.
In contrast, earlier yDNA
testing was of STRs. A STR
("stir"?) is a short tandem repeating sequence of bases defined by its
pattern of As, Cs, Gs and/or Ts; some STRs are found in multiple positions.
A test for a STR marker results in a number counting how many times the pattern
SNPs are named somewhat peculiarly. Each name consists of two parts; firsat is
one or more letters, a code for the discovering group. The letter or letters are
followed by a number represetiong the order in which the SNP was found by that group.
A SNP may be discovered and named independently by more than one group. It will then
get two or more names, all referring to the same thing. A standard
convention in listing these is to give all names, separated by a slash,
There are also "equivalent" SNPs. These are SNPs which are always found together; if
one is present, so are its equivalents. The convention for these is to
separate them by commas, e./g. "Z147, CTS6554, S489, Z29612".
Because the research process is to first identify SNPs, then determine
their places on the phylogenetic tree, SNP names have little relationship
to the tree structure.
Why test ySNPs?
Unlike STRs. SNPs accumulate; old ones don't go away but new ones get
added. SNPs do not mutate back to earlier forms. Thus, they retain a
permanent record of the chromosome's mutation history over thousands of
generations. The younger the SNP, the more specific the information about
Another advantage of ySNP testing is that results stand on their own. Unlike STRs, you don’t really need others
in your patriline to go along. Results can place one on a branch of the human phylogenetic tree,
depending on the state of knowledge about the branch and its progenitors.
This knowledge presently has many gaps; the gaps are being gradually
filled, as more people test an provide more . While your SNP results aren't
compared to a specific person, they are compared to models derived from
Some ySNPs have been found which are unique to particular genetic families. These ySNPs
are definitive for belonging to their patrilines. (For example, L555 is
definitive for the Clan Irwin "Borders" family.) At present, only a few of
these have been found and verified.
Upgrading to more STR markers (say, 67 or 111) could help to increase
confidence in and precision of STR matches – if there are others to compare to. More markers may (with enough results) help to define the
family's branches. So far as branch identification, our project's experience hasn’t been so good; we seem to
be left with too many remaining ambiguities.
We recommend either (not both) of two strategies:
- NGS testing route
- SNP Pack route
A prior ySTR test is recommended or required) before either NGS or SNP Pack tesitng.
NGS (Next-Generation Sequencing)
Either FTDNA’s Big Y (~$575) or Full Genomes Corporation’s Elite ($775)
are like drinking from a fire-hose. They are the “Gold Standards” of ySNP testing at the moment. They give you lots of information,
>11,000,000 base pairs, more than anyone has, so far,
really made sense of.
We recommend that folks who choose the NGS route follow up by buying an interpretation of the results from YFull,
a separate company.
Less comprehensive, but more comprehensible, are SNP Packs available from
FTDNA and YSEQ. Their costs run from $99 to ~$120 . They involve a "zeroing-in" process. Test more generally first, then
more specifically based on the prior results.
If, for example, predicted to be R-M269, start with the R1b-M343 Backbone Panel ($99); it's
a bundle of 142 ySNPs which will confirm (in most instances) the R-M269 prediction and define the haplogroup to about six levels finer.
That first panel will point you toward the next SNP Pack to order; another 100+ SNPs for ~$99;
it will drill down another 4-5 levels
of haplogroup delineation, A third SNP pack (same numbers & cost) is about as
far as one can go now.
Testing individual SNPs, one at a time, is inefficient; it is an expensive and
minimally informative approach.
It might seem like NGS is better because it’s one shot and perhaps no more expensive than a series of SNP Packs. The problem is that making sense of the NGS results could take years of effort. With SNP Packs, you’re getting information which has (mostly) been investigated and catalogued.
The problem with SNP Packs is that (with few exceptions) they won't find
"private" ySNPs to identify a particular patriline.
With either NGS or SNP Packs, you’ll see many negative ySNP results but focus on the positive ones.
The SNPs which turn out to be negative were included for elimination purposes
and to "cover all the bases". Negative tells you what you aren’t; positive tells you what you are.
Example: The R1b Backbone Panel includes a L21+ result. You will also
see positive for P312+, P310+ and M269 -- successively larger groups which
include L21+. You will also see U106-, L47- and many others.
Clearly, there are many more things anyone is not than what anyone is. The
same is true for ySNPs.
With NGS testing, there will also be "no calls" and other hard-to-decipher
findings. Interpreting these is not for the faint of heart.