Today we would like to explain the several types of ancestry/admixture analysis available:
Firstly, it is important to note that Admixture and Ancestry are distinct concepts.
Ancestry refers to your actual and recent ethnicity and attempts to answer the question “What is my genetic ethnicity composition?” For example, someone who is 100% Ashkenazi Jewish has a genetic ethnicity composition of 100% Ashkenazi Jewish.
Conversely, Admixture breaks down your genetic composition to determine the best-fit populations that match your DNA. It attempts to answer the question, “What is the composition of my ethnicity?” For example, someone who is 100% Ashkenazi Jewish has an admixture of 50-60% Levantine, 30-40% Southeast Mediterranean, and a minor percentage of East-Central European.
LAI (Local Ancestry Inference): Segment-Chromosomal level
The most advanced and recent technology for ancestry analysis.
This kind of analysis divides your chromosomes into segments of 200-500 SNPs (depending on the company). Infers the most likely population for each segment and smooth the results to show the longest most likely population for each chromosome and side.
Useful for: Recent ancestry (last 1.000 years).
Companies that offer this type of analysis include:
- DNAGenics: Shared Modern Origins and Deep Ancient and Modern reports
- 23andme
- AncestryDNA
- FTDNA (since sept. 2021)
- Genomelink
Admixture analysis
An admixture analysis is a type of genetic analysis that breaks down an individual’s DNA into the best-fit populations that match their genetic makeup. It attempts to determine the different ancestral populations that have contributed to an individual’s genetic composition. Admixture analysis can be helpful for both ancient and modern ancestry research, but it cannot tell an individual’s recent ancestry. The results of an admixture analysis can provide insights into an individual’s ethnic and racial origins, migration patterns, and historical population movements.
Useful for: Ancient and Modern admixture analysis
Companies that offer this type of analysis include:
- DNAGenics (free and exclusive calculators)
- LivingDNA
- Gedmatch calculators
- YourDNAPortal (free calculators)
- MyHeritage
- Tellmegen
Oracle analysis
This type of analysis performs a mathematical comparison of your admixture or G25 Coordinates results against a set of populations. The best fit using a Least-Squares algorithm is the sample that is mathematically closest to your results. This is one of the oldest methods used to find the closest population. It is accurate for non-mixed people.
Oracles are limited in their ability to determine mixed ancestries, except in cases where the mix is at the continental level (such as a mix of African and European or Asian and African ancestry). However, at the subcontinental level, Oracles cannot accurately determine the ancestral breakdown.
So, if you have no mixed ancestry, the analysis results will fit well enough.
Useful for: Non-mixed individual
Companies that offer this type of analysis include:
- DNAGenics: Admixture Studio > Population Studio
- Gedmatch calculators + Oracle
- YourDNAPortal paid reports: Speculative and Conservative modes of Pangea, African ancestry, Native American ancestry, some of their free admixture calculators.
Montecarlo analysis
This analysis tries to find the best combination of N populations that best fit your Admixture / G25 Coordinates results. This analysis is good for finding minor ancestries and ancient ancestry.
Useful for: Minor ancestries and ancient ancestry
Companies that offer this type of analysis include:
- DNAGenics: Admixture Studio > Population Studio. We offer a MultiEthnic algorithm that improves the results for mixed people.
- YourDNAPortal: paid reports Pangea, African ancestry, Native American ancestry
** Extra: G25 Coordinates (Oracle/Montecarlo)
G25 Coordinates are PCA coordinates (Principal Components Analysis) calculated using SmartPCA or PLINK. These tools calculate the “best statistical distribution” of a set of populations and your genome. You can then use these coordinates to compare them using Oracle/Montecarlo tools.
