Different testing laboratories use different calibations. As a result, it is necessary to make conversions of the data from one laboratory before mixing it with data from another laborarory. This issue has become more complicated than ever before due to changes made by Ancestry.com in preparation for their takeover of Relative Genetics. It has been difficult to keep up with all of the changes. The best and most comprehensive discussion that I have seen is the following message posted to the DNA Forum at Genealogy.com by Robert Stafford: http://genforum.genealogy.com/dna/messages/1945.html.
Note that the need to convert DYS452 and DYS463 from FTDNA to YSearch has been eliminated as of 1-Mar-2010, when FTDNA realligned the data in YSearch to bring it into compliance with the NIST standard (i.e. the same as used by RG and SMGF)
Since I have taken the liberity of extracting data from other databases and adding them to my project, I have made every effort to properly account for the necessary conversions. For example, the following conversions from/to the Family Tree DNA (FTDNA) / Y-Search standard to/from the current Relative Genetics (RG) / DNA Ancestry standard are necessary:
These conversions are appropriate for the current Relative Genetics / DNA Ancestry standards, including the recalibrations done in August 2007. Some web sites, Y-Search and www.ybase.org for example, are still using the old Relative Genetics standards in place before August 2007 and referring to it as "the Relative Genetics Standard". The user should be cautious on this issue
If the data are from DNA Heritage or www.ybase.org prior to December 2007, the calibration standards are identical to the old Relative Genetics standard (i.e. before the recalibration done in August 2007). This standard was also identical to that used by FTDNA and Y-Search except for GATA-H4, wihch was reported as TAGA-H4. Use the following rule to convert from/to the old DNA-Heritage/www.ybase.org value for TAGA-H4 to/from the current Relative Genetics (RG) / DNA Ancestry value for GATA-H4.1:
In December 2007, DNA Heritage adopted the new Reletive Genetics standard and converted all of their data. Unfortunanely, they chose not to convert the data on www.ybase.org leaving that task as the responsibility of whomever uploaded the data to the web site in the first place. This has created the obviously unacceptable situation of having a mixture of data from the old standard mixed with data from the new standard. What could they be thinking? It is obvioulsy unreasonable to expect that people who uploaded data in years gone by will come back and revise that data. For one thing, some of them are undoubably dead by now.
Clearly the people running these testing companies and data repositories don't have a clue as to the meaning of a standard. In my humble opinion, the only thing we can do to deal with this situation is to simply ignor the value of any marker that has been contaminated by this irresponsible redefination.
Allow me to explain the concept of a standard. Take for example the foot, which is 12 inches. If there where a reason to redefine the foot to be 13 inches, it would be completely irresponsible to just simply decide that starting today, the foot is now 13 inches. If we decided to start using a 13 inch foot, we would be wise to give it a new name, such as "new foot" or "big foot". As an example, the change of value for GATA-H4 accompanied by the change of the name to GATA-H4.1 is a reasonable and acceptible change.