Research

Rox2 was identified as a monophyletic haplogroup and was named (through STR analysis) a few years before NGS testing identified DF27 and its subclades.

• In order to get maximum information and value out of  expensive Big Y or FGC tests it is important to compare with other results on public phylogenetic trees.  Most of the initial DF27>ZZ12 discoveries, i.e. those made before 2019, have been made at the Big Tree.  If your particular subclade/branch below Rox2 currently has only a few members, suggest Big Y/NGS testing to your closest STR matches - they might help populate your branch.  There are occasional Sales at FTDNA for Big Y, usually at Christmas.

• Know the subject, your goals and what to expect before testing and choose the type of test that can help achieve those goals.  This will help to avoid wasting money.  DNA genealogy does not give all the answers in one test - for best results it requires as much active research as traditional genealogy does.  New test results lead to more questions.  Trace your paper trail as far back as possible in Europe and bear in mind that surnames can swiftly change but yDNA does not.  

• My motivation stems from an interest in family history and local history.  Trust your own judgement - details are far from certain.  New evidence from archaeogenetics can potentially change the claims made in DNA projects overnight.  Scientific objectivity is key, maintain an open mind and be prepared for surprises.  This website presents my interpretation of the currently available data, others might interpret the same data in a different way.  Different age estimates can result from different methodologies and data.  Firm answers will come with more evidence, especially more tangible radiocarbon dated results from ancient yDNA studies/archaeogenetics.  'The directional nature of time means the past generally explains the present and not vice versa.' DMXX

• The relatively high number of close STR matches with different surnames is normal for our subclade. Rox2 appears to have been prolific and produced many sons in medieval times and the resulting STR matches are not always due to recent NPEs - they could be due to lack of STR divergence, as mentioned on the homepage.  In most cases the yDNA connection with other Rox2 matches is many centuries ago - often before surnames started to become fixed in around the fifteenth century.  Only Big Y, not STR matches, can tell for sure if you have found a recent cousin or a more distant one. 

• Both SNPs and STRs regularly vary from statistical norms.  SNPs don't happen like clockwork in a predictable chronological manner.  Generic online SNP estimates (years-per-SNP) are based on the average for everyone, not individual idiosyncratic subclades.  Rox2 has several different branches with their own unique average number of SNPs.

• Ensure privacy settings at FTDNA are set to allow matches to find you.  Matches span a number of surname projects with origins in different countries and it is always interesting to find a new fellow Rox2 match.  Be cautious about giving out personal information on social media sites (e.g. Facebook or Twitter).  To share STR information with fellow FTDNA customers/genealogists, sign in to your FTDNA account and at the top right of the FTDNA 'myDashboard'.  Click your name, then click Privacy & Sharing next to the padlock symbol and select 'All Levels' for Y-DNA Matching.  Check 'Opt in' for Origin Sharing and check 'Opt in' for your Project Sharing/Group Project Profile.  This does not reveal personal details, it just helps others to compare their STR markers with those in the database.  

• In May 2018 EU legislation on data privacy (General Data Protection Regulation, GDPR) was adopted by FTDNA.  The customer must balance any privacy concerns with genealogy goals.  Excessive privacy limits the value and purpose of the DNA test - it's less likely that you will learn anything.

• Nodes/branches on the phylogentic tree, made from modern, self-reported hobbyist data are not conclusive proof of geographic origins in prehistory.  Ancient yDNA results, when compared with modern phylogenetic trees, will provide the best evidence.

• Consider creating a website or a blog for general genealogy information (e.g. WordPress, Google Sites or Blogger).  It is interesting for matches to follow progress in different parts of the Rox2 tree.  I update and edit all the pages on this website with new information when it comes to light, so it's worth bookmarking it and returning occasionally for a recap.

RESEARCH HISTORY

The cluster was named 'Rox2' by Jim Turner.  Jim and I tested in 2005 and found we had closely matching STR markers along with an off-modal 'signature' pattern shared with a few other kits in the early, sparsely-populated yDNA databases.  DNA genealogy kits at that time produced only 12, 25 or 37 STR markers.  The term Rox2 has stuck after initial discussions in 2006 - it was, and is, a useful shorthand way to describe a group known by a complex signature STR pattern and several different equivalent alphanumeric SNP names.

Continued STR research between the years 2005 and 2012 indicated (to some of us) that Rox2 was a monophyletic clade (see Estimated Age page) but the deep R1b-P312 subclade it descends from would remain unknown for some time.  Many did not, and still do not, recognize the existence of monophyletic STR clusters without SNP proof.  That proof took longer to arrive than it did for most similarly-sized groups under U152, L21 and U106.  L21 was 'discovered' in 2008 but DF27 remained invisible  and unknown within the undefined R1b-P312* paragroup until 2012.

Rox2 matches tested negative for L21 soon after its discovery and were also negative for all SNP discoveries below P312 that were made available for 'stand alone testing' between 2008 and 2012.  Over that time STR resolution increased to 67 and then 111 markers and the extra definition meant that Rox2 matches could be identified with more certainty.  25 and 37 STR marker matches upgraded and several more signature off-modal STR markers were revealed in the new panels.  At that time, unlike now, the entire database could be accessed by ordinary customers/hobbyists.  This website was created in 2009 in the absence named SNPs and a dedicated FTDNA project.  Proud Scot, Angus Horatio Stewart, kit 142928, was the first Rox2 STR cluster match to test positive (derived) for DF27 when that SNP was identified and became available to test in May 2012.  This proved beyond doubt that Rox2 is a monophyletic clade and vindicated the conclusions arrived at through STR analysis alone.  The whole genome of a 50000-100,000 year old Neanderthal was successfully sequenced by the Max Planck Institute in 2013.  NGS testing by Rox2 matches revealed the complex structure of Rox2's phylogenetic tree after 2014.  

Chip-based SNP tests were not able to see most of what would later be identified as the large ZZ12 section of DF27.  Those tests were not even able to identify the DF27 SNP itself.  So, from 2012 until March 2014, Rox2 remained a R1b-DF27** subclade (the double asterisks mean that Rox2 was negative for all then-known SNPs below DF27).  Parallel branch DF27>Z195/Z196 was identified long before DF27 or ZZ12, it being easily-identified on the first chip-based SNP tests.  Despite members of ZZ12 subclades taking chip tests like Geno 2.0, 23andMe and Chromo2, no SNPs below DF27 were identified.  However, it was clear (to some of us) that there was more to DF27 than just Z195.  Of the 144 DF27 kits in an early Chromo2 spreadsheet (2014), 61 were DF27*, i.e. 61 were positive for DF27 and negative for all other then-known subclades.  These unsatisfactory results clearly indicated that many ZZ12 subclades were being missed but the STR evidence of their existence was not generally accepted until the SNP 'proof' arrived in 2014.  Even 2019 tests, like LivingDNA, using the Illumina GSA chip, could not read DF27.  Less expensive chip tests (1240K panel) used to analyse ancient remains only produce a R1b-L151 result for DF27>ZZ12 subclades.  ISOGG 2019 still had patchy coverage of DF27>ZZ12.  The first proper identification of the whole of DF27 happened at the Big Tree  after analysis and inclusion of NGS results belonging to members of DF27>ZZ12 subclades by Alex Williamson in 2014

Turner (FTDNA kit N3036, FGC kit number GYBGZ) received FGC results on 3rd April 2014 (link).  Many SNPs were seen to be shared with an anonymous kit in the 1000 Genomes project, HG00107.  HG00107 had Orkney ancestry and matched the Rox2 off-modal STR pattern.  Dickinson (FTDNA kit 134765, FGC kit WBAFF) received FGC results on 12th April 2014.  Those two pioneering tests highlighted the  dozens of shared SNPs in a phylogenetically equivalent block below DF27>Z2571.   FTDNA's first incarnation of Big Y (BigY1 and BigY2) read less than half of the 45-ish Rox2 equivalent SNP block present in the two FGC tests but 'Private' SNPs below the shared block compared well.  As the Big Tree began to fill out with new Big Y Rox2 matches after 2014, the number of named parallel 'son' subclades steadily increased.

With increasing commercialization and data privacy legislation, the 'golden age' for genetic genealogy began to fade after 2015.  In March 2015 FTDNA made new customers' results private unless that customer went on to change their settings manually to enable sharing.  Existing customers were no longer able to see their new matches' STR results unless they were a members of the same project.  There were once several different, searchable online yDNA databases but they have disappeared one by one since 2015.  The Sorenson (SMGF) database was an excellent free resource for finding matches that had 100,000 kits with attached family trees but it was taken offline in 2015.  

In summer 2017, FTDNA's SNP coverage of DF27 greatly improved with the inclusion of the Big Tree's new ZZ12 discoveries on its Haplotree.  FTDNA began to give 'BY' (Big Y) names to new SNPs.  The three years of SNP discovery on the Big Tree meant that by August 2017 there were more branches and children below DF27>ZZ12  represented there than any other subclade of R1b-P312.  As of March 2018 there were twenty-three 'sons' of DF27>ZZ12 and several more waiting to be discovered.  One year later there were 41 on the Big Tree and numbers continue to rise as new branches emerge.  In August 2019 there were 29 parallel branches immediately downstream of DF27>ZZ12 on the FTDNA Block Tree plus many more ZZ12* kits that have not yet found their named deep subclade.  By March 2021 there were 34 'sons' of ZZ12 plus the ZZ12* paragroup.

Further data privacy laws introduced in 2018 again reduced the visibility of matches.  It became more difficult for FTDNA customers to compare their data with each other for genealogical purposes.  YSearch was one of the most useful, large STR databases accessed from FTDNA as part of their original STR testing package.  It was possible to find matches beyond genetic difference of 7 over 67 markers.  YSearch was used to get in contact with many Rox2 matches, including Stig (N3285), the first Swedish Rox2 STR match, in 2013.  In addition, genetic genealogists could upload experimental and modal haplotypes to the YSearch database as well as ones from various different companies and external archaeological studies.  YSearch was eventually taken offline in 2018.  

In 2018 FTDNA upgraded Big Y results from build hg19 (2009) to hg38 (2013).  In late-2018/early-2019 FTDNA released the Block Tree as their copy of the Big Tree and also rolled out Big Y-700 to replace Big Y-500.  Big Y-700 (BigY3) now matched the 2014 FGC Elite coverage.  Unfortunately, unlike its inspiration the Big Tree, FTDNA's Block Tree does not display kit numbers or names for all the Rox2 Big Y matches displayed on it.  See FTDNA's Big Y-700 White Paper.   

New SNPs from Big Y-700 and as-yet-unfiltered 'bogus' and/or recurrent SNPs mean that the 30 SNP mismatch limit originally set for Big Y-500 tests was being exceeded for many Rox2 kits.  Matches can remain invisible as a result - the SNP names and branches will appear on the FTDNA Block Tree but there are no surnames or kit numbers to help us to identify them.  Some kits also have high genetic distance between their STR results and therefore exceed the STR matching threshold and do not show up as STR matches either.  There is no easy way of knowing if the several anonymous kits below the FGC11397* equivalent block on the Big Y Block Tree have tested with Big Y or have just tested for one SNP, i.e. FGC11397.

It is relatively straightforward for a potential Rox2 match to take a NGS test and be quickly moved to the correct part of the phylogenetic tree.  NGS tests and FTDNA's Block Tree have simplified list-making.  Accurate interpretation still requires experience and a phylogenetic tree with identifiable surnames.  NGS/Big Y defined haplogroups are now well-represented on the SNP-led phylogenetic trees and large, young NGS-defined subclades have become hard to miss.  The NGS kits currently displayed on the Big Tree, YFull's tree and Block Tree are the tip of the iceberg and are the early pioneers of DNA genealogy.  There are many Rox2 matches scattered throughout hobbyist databases that have not yet taken NGS test or who may have taken an early SNP chip test with no result.

New (2022) T2T analysis is being applied to results. The Human Genome project that appeared in 2003 used technology that left areas of the y chromosome unmapped.  T2T is expected to address these issues - issues that led to large parts of the DF27 haplogroup being unrecognized by yDNA testing.  The new SNPs being discovered are labelled 'FTT' on FTDNA's Haplotree.  It is possible that new SNPs will be discovered for Rox2 - ones that might change the structure and branching of the current phylogentic tree.

In July 2022 FTDNA launched a Beta version of their new Discover tool.  It is a clear and easily-navigable way to understand the structure of Rox2 branching with accompanying age estimates.  Further updates over the following year included Globetrekker in July 2023, a map displaying estimates of  'geographical ancestor locations and migrations'.  Bear in mind that such maps require accurate self-reported places of birth in Europe for the sample's earliest known ancestor and are affected by bias in the customer databases (DNA for genealogy samples are not drawn from an even distribution across all territories but reflect areas of greatest migration to the New World).

Rox2's tree has the potential to grow if new results continue to be shared and the price of testing falls.  However, new yDNA matches now show up less frequently than they used to.  New Rox2 yDNA matches have slowed from a steady flow before 2015 to a slow trickle by 2020 and to an intermittent drip in 2023. The trend within DNA genealogy is moving towards privacy and as a result it is becoming difficult for customers to compare their own data on increasingly anonymous trees.  DNA genealogy as a hobby works well if customers can easily compare and communicate their results with each other - with the reassurance that those results are not being misused outside the hobby.