Recently Roots for Real has been mentioned by a webpage "Reviews for Roots for Real" which confusingly does not rank by accuracy of the test results1. The correct ranking would be:
- Rank 1
- Roots for Real2 (jointly with four other companies A3, B4, C5 and D5)
- Rank 6
- BritainsDNA 6
- Rank 7
- Other company
- Rank 8
- Other company
Although we do well in terms of accuracy, we caution that any useful survey would need to distinguish between autosomal tests, and sex-specific tests (mtDNA and Y tests). All autosomal tests, including our own, are biologically "fuzzy", because they trace DNA from many ancestors and try to assign them to populations/races/nations which generally do not have sharp boundaries. The uncertainty in autosomal DNA even allows scholarly debate whether or not we have one percent of Neanderthal ancestry. mtDNA and Y tests are much more precise (but at the cost of tracing only one lineage in each case). A major reason for performing autosomal tests anyway is that some women do not have access to male family members to trace their father's ancestry.
So how does one choose a suitable company and a suitable test? There are in my view three key points: 1. The quality of DNA markers included in the test. 2. The size of the comparative database. 3. The quality of the database. The third point is arguably the most important, but also the most difficult for a consumer seeking a truthful answer. Ultimately it is a matter of trust, like all else in life.
Peter Forster, Cambridge, UK.
1. The corrected ranking is based strictly on the percentage non-white (non-European/non-Middle Eastern) admixture estimates reported for their English test subject, taken from the DNAtestingchoice website on 2014 March 3. Their website may change, so we have documented their original reports in the following footnotes.
2. Roots for Real. "[The map] showed that 100 percent of my DNA was from Europe"
3. Company A. "Going back to the results I received in the post, the ‘certificate of authenticity’ contained a single ‘population percentage’ bar chart showing I shared 93.46% of my DNA with populations from ‘Western Europe – Orcadian’, and 6.54% from ‘the Middle East – Palestinian, Bedouin, Jewish and Mozabite’. Both results were presented with a 1.4% margin of error which meant the accuracy of these results was high. I was surprised that the main population group I had the strongest match with was Orcadian – my family had no knowledge of any ancestors from the Orkney Islands!"
4. Company B. "99.7% European, of this I was 38.3% British & Irish, and 54.3% non-specific Northern European"
5. Company C and D. "Graphs were shown that clearly indicated my DNA was matched to Northern and Southern European populations, and to European immigrants. [...] My top personal ‘regional population’ matches were with Southern European populations, Northern European and with European immigrants. I was interested to see that I was more closely matched with Southern European populations than with Northern European populations.
6. BritainsDNA. "69% European, 13% West Eurasian, 9% Southwest Asian and 4% Pan-African – the remaining percentage was apparently ‘statistical noise’"