In order to understand the difference between a confirmed and a predicted haplogroup, it is important to first understand the difference between haplotypes and haplogroups. You may have heard of the terms haplotype and haplogroup, and while these two terms sound very similar, they have different meanings.
Haplotypes
Each person’s DNA is unique in much the same way that each individual’s written signature is unique. An individual's unique DNA sequence is called a haplotype, also sometimes referred to as a DNA signature or genetic signature.
A haplotype can refer to the entire genome of an individual, but can also refer to any portion of the genome, all the way down to a single nucleotide. This is useful for genealogists who only want to focus on small portions of DNA that are useful for matching or research. Regardless of the size of the portion, the genetic signature found within the segment(s) you have selected is the haplotype for that segment. To better understand this concept, let’s use the idea of a name as an analogy.
Let’s say your name is John Doe, and you want to find your name in a database. If you only type the letter J into a search, you will likely find many people who all have a J in their name. To help narrow it down, you next search for JO. There will be fewer people, but Joel, Jordan, Joseph, and many others will still appear. Next, you type in JOHN, and the search narrows it down even more. The complete collection of all of the letters in your name (J O H N D O E) is unlikely to be shared by many people in the database. In other words, the more detailed you make your search criteria, the fewer people will match the criteria.
Now let’s think of this name as your haplotype. If you define your haplotype as the complete JOHN DOE, you will probably be the only person in the database with that haplotype. Instead, if you define your haplotype as JOHN, you will now share that haplotype with other people. Regardless of the size of the haplotype you define, all of the people who share the same haplotype are grouped together to form a haplogroup.
While this analogy hopefully helps you understand a haplotype, in reality, we use Y-DNA STR marker values, not letters, to define a haplotype. You can read more about Y-STRs here.
Haplogroups
Continuing our name analogy, let’s think about the entire alphabet. If your haplogroup is simply defined as everyone whose name starts with J, then it will exclude the majority of possible names, but your haplogroup will still have many, many names within it.
You can narrow this haplogroup to JOHN. Everyone in haplogroup JOHN will also be in haplogroup J, but not everyone in Haplogroup J will be in Haplogroup JOHN. By using this method, we can assign large, broad haplogroups with many smaller haplogroups inside them. For example, JOHN would be a subgroup of JO, which in turn would be a subgroup of J. We could write out this haplogroup path as J > JO > JOH > JOHN.
When we map out all possible combinations of letters (or STRs, in this case) we get a series of connected haplogroups that looks very similar to a genealogical family tree. With some haplogroups having many descending branches, while others have few.
Neighboring branches will have similar, but not identical haplotypes. They will also share many identical STR values. This is one reason STRs are not ideal markers to define our tree. Instead, we use Y-DNA SNPs (pronounced snips). You can read more about Y-SNPs here.
Predicted vs. Confirmed Haplogroups
Each major Y-DNA haplogroup has many, many subgroups within it, each defined by a different Y-DNA SNP. Members of neighboring branches tend to have similar STR haplotypes. This means we can look only at the Y-STR haplotype of an individual, and be able to make a reliable prediction which branch they belong to, even if they have not tested for the SNP that defines that branch. If the test taker does not have a SNP confirmed haplogroup, the instead receive a haplogroup prediction.
Because neighboring branches can be very similar, and share STR marker values, we pool these values together to form a range for each parent haplogroup. This haplogroup is generally far back in time to include possible variations. The prediction we make based on haplotype will only correspond to a very ancient haplogroup for this reason. In order to confirm to which subgroup a person belongs, SNP testing is required.
In some cases, we are unable to accurately predict a haplogroup if these STR marker values do not fall within a typical range. In cases such as these, FamilyTreeDNA runs a complementary test for defining SNPs called our SNP Assurance Program, sometimes called a Backbone test. This collection of SNPs, called a panel, includes the defining SNP for each ancestral haplogroup in order to accurately provide an ancestral haplogroup for customers. You can read more about the Backbone test here.