Ancestrify

Ancestry populations · Classical Antiquity

Numidian Berber

The Berber-speaking population of Roman North Africa — Numidia and Mauretania — before the Arab conquest. One of the 35 curated source populations in the Classical Antiquity era of the Ancestrify qpAdm ancient-DNA analysis.

See plans and pricingBrowse every population

Who they were

Artistic interpretation of the Numidian Berber (AI-generated artwork, not documentary evidence)
Period
0 – 500 AD
Era in the analysis
Classical Antiquity
Region
Numidia and Mauretania, Roman North Africa
Role in the model
Source (left) population — a reference qpAdm tests against

The Numidians were the Berber-speaking people of the Maghreb interior who fought alongside and against Carthage and Rome, and whose kingdoms became the Roman provinces of Numidia and Mauretania. This source represents the indigenous population of Roman North Africa in the centuries before the Arab conquest, heirs to the deep Iberomaurusian and Neolithic Maghrebi lineage with a modest Punic, Iberian and Roman-era Mediterranean admixture along the coast (Fregel et al. 2018; Serra-Vidal et al. 2019).

This ancestry is the pre-Islamic foundation of present-day Algerians, Tunisians and Moroccans, and it appears in Iberia, Sicily and southern Italy through Roman-era and medieval contact. In the model it is the earlier of the two Berber sources: it differs from the Indigenous Berber source in lacking the Arabian and sub-Saharan contributions of the medieval centuries, and from the Carthaginian source in its lower Aegean- and Levantine-related share.

In Ancestrify's qpAdm service, Numidian Berber is a source in the "Classical Antiquity" era. The report shows its weight with a standard error and a Z-score, so you can see not only how much of the model it carries but how confidently that share is separated from zero, and a model that cannot support it is rejected rather than published. For genomes from the Maghreb its weight against the Indigenous Berber source describes how much the model attributes to the pre-Islamic population rather than to the later admixed one, and the two related references may be hard to separate.

Y haplogroups E-M81 and E-M78, and mtDNA U6, M1, H1 and T2, are characteristic of these populations. A haplogroup alone is never evidence of a share.

In your report

The Numidian Berbers of Roman-period Algeria (~146 BCE–430s CE), represented by the Numido-Roman Berber genomes (Algeria_NumidoRoman_Berber.SG) — indigenous Berber (Amazigh) peoples shaped by millennia of local genetic continuity and successive Mediterranean contacts. Ancient DNA from the broader Maghreb shows that indigenous North African lineages trace back over 20,000 years to Iberomaurusian foragers, forming a deeply rooted genetic substrate. The Neolithic transition introduced European and Levantine farmer ancestry, and subsequent Phoenician colonization (from ~1200 BCE) added further Mediterranean diversity before Roman conquest. Under Roman rule, North Africa became one of the empire's most prosperous regions, with cosmopolitan cities like Carthage, Leptis Magna, and Hippo Regius facilitating continued demographic exchange across the Mediterranean. Later Vandal, Byzantine, and Arab demographic inputs further transformed the region's genetic landscape, though Berber-related ancestry remains prominent in modern Maghrebi populations.

Artwork is an AI-generated artistic interpretation informed by the archaeological record — not documentary evidence.

Where this population sits in the model

The Ancestrify qpAdm analysis models your genome once per era. The Classical Antiquity era has 35 curated source populations, and Numidian Berber is one of them: every candidate model for that era is a combination of these sources, tested against a fixed set of outgroup populations, and only a model that passes the statistical fit check is published — with a p-value, and a standard error on every proportion.

A reference, not a verdict: every population on this page is a source the model tests your genome against — a well-sampled point in ancient genetic space that a mixture of ancestries can be expressed in terms of. A weight on it means your genome is well described as partly resembling those people; it is never a statement that they, specifically, were your ancestors, and a population that fits well is not the only one that could.

Read how the method works on the qpAdm analysis page, or the definitions in the glossary.

Model it yourself for free

The Ancestrify Lab runs in your browser, with no account. Paste a Global25 row into an admixture calculator that covers the Classical Antiquity era, rank the closest ancient populations in the distance tool, or run f4, qpWave and qpAdm on the public panel in the AdmixTools 2 workbench. These are Global25 fits and exploratory statistics, not the formal qpAdm model of the paid analysis.

Questions

  • Who were the Numidian Berber (0 – 500 AD)?

    Numidian Berber is one of the 35 curated source populations in the Classical Antiquity era of the Ancestrify qpAdm analysis: a group of published ancient genomes from one archaeological context, pooled as a single reference the model can test your genome against.

  • Does Numidian Berber in my result mean they were my ancestors?

    No. A source population is a reference the model tests against, never a statement about who your ancestors were. A coefficient on this source means the ancestry those burials represent helps explain your genome to within statistical noise — similar ancestry composition, not descent from those individuals. Ancestrify never claims a population is your ancestor.

  • What does a Numidian Berber percentage in a qpAdm model mean?

    qpAdm estimates the mixture proportions that best explain your genome as a combination of the chosen sources, given a fixed set of outgroup (right) populations, and reports a p-value for whether the model is statistically acceptable together with a standard error for every proportion. The percentage is that estimated share, with its uncertainty published beside it.

  • How do I get a qpAdm model that uses Numidian Berber?

    Order the Ancestrify qpAdm analysis (€29.99, one-time) and upload the raw file you already have from a consumer test. Your genome is merged into the AADR reference panel and modelled across both eras; each era's model is built from that era's source populations, so a model for the Classical Antiquity era draws on Numidian Berber where it improves the fit.

  • Can I model Numidian Berber ancestry for free first?

    Yes. The Ancestrify Lab runs Global25 admixture calculators, a nearest-populations distance tool and an AdmixTools 2 workbench in your browser with no account, and several calculators cover the Classical Antiquity era. Those are G25 fits, not formal qpAdm models — the paid analysis is the one with a p-value.

More populations from the Classical Antiquity era

Browse the full directory of source populations, or continue with another population from the same era:

Ancestrify

Combining cutting-edge genomic science with rich historical records to map your ancestry across generations and continents.

© 2026 Ancestrify. All rights reserved. · Ancestrify is a trading name of Andi Thomaj, a sole trader registered in Tiranë, Albania · NUIS M61725001N
Card payments processed by POK Payments (RPay Ltd)VISAMASTERCARD