Bangladeshi DNA Ancestry: What Your 23andMe, AncestryDNA or MyHeritage Result Means

If your family is from Bangladesh and you grew up in Tower Hamlets, Birmingham, New York, Toronto, Dubai or Sydney, there is a good chance someone in the family has already taken a DNA test. There is an equally good chance the answer came back as one large block of colour: Bengali & Northeast Indian, Bangladesh, Bengali, or the even less helpful Broadly South Asian. It confirms what you already knew and explains none of it.

This guide explains what those labels are built from, what published genetics says about people from Bangladesh, and where the honest limits are for Sylheti, Chittagonian and other regional backgrounds. For the wider Bengali picture, including West Bengal, see our guide to Bengali DNA ancestry; to trace a particular village, see finding an ancestral village in Pakistan or Bangladesh with DNA. This post is about the Bangladeshi result already in your account, and what you can do with it.

The labels you were given, decoded

Each company groups its reference samples in its own way, and both 23andMe and MyHeritage have split their South Asian categories since 2024. The label on your screen therefore depends on which company you used and when you last looked.

Company	What you may see	What sits behind it
23andMe	Bengali & Northeast Indian, or Bangladeshi & Northeast Indian; sometimes Broadly South Asian; possibly a Bengal Basin genetic group	In July 2024, 23andMe listed this population as built from Assamese and Bangladeshi reference samples and called it one of its broader clusters. In January 2025 it added 44 finer South Asian genetic groups, the Bengal Basin among them.
AncestryDNA	A Bengal region, often with smaller shares in other South Asian regions	A region drawn around Bangladesh. The percentage measures similarity to Ancestry's reference samples, not a count of Bangladeshi ancestors.
MyHeritage	Bengali, or the older broad South Asian	The current ethnicity estimate lists Bengali separately from South Asian and from Pakistani and Punjabi. Older reports may show only the broad label.
FamilyTreeDNA, Living DNA	Usually one or a few broad South Asian regions	Labels differ by version, so check your own report. Both companies let you download raw data, which is what matters for re-analysis.

None of these labels tells you which part of Bangladesh your family comes from, or how your ancestry divides into the deeper components geneticists use. If most of your result sits in an unassigned bucket, our explainer on why 23andMe says Broadly South Asian covers what that means.

Why Bangladeshis often score more "Bengali" than West Bengalis

A cousin from Dhaka or Sylhet often gets a higher Bengali & Northeast Indian figure than a friend whose family is from Kolkata. That says nothing about who is more Bengali. Two technical things produce it.

• The reference samples. When a category is built from Assamese and Bangladeshi samples, the people who most resemble those samples score highest. 23andMe has said its references include 1000 Genomes Project samples, and that project's South Asian panels include one labelled Bengali in Bangladesh.
• The eastern component. The most distinctive feature of Bangladeshi genomes, compared with the rest of South Asia, is a minority share of East Asian-related ancestry. A classifier separating a Bengal category from a North Indian one leans heavily on it: a little more pulls you east, a little less pulls you west.

So the number is mostly a reading of one feature of your genome. It is not a measure of language, culture, faith or belonging, and DNA does not decide who is Bangladeshi or Bengali.

What published genetics shows about people from Bangladesh

The same deep layers as the rest of South Asia

Reich and colleagues showed in 2009 that most groups in the subcontinent descend from a mixture of two ancestral populations, Ancestral North Indians and Ancestral South Indians (see ANI and ASI ancestry). Moorjani and colleagues dated that mixture in 73 groups to about 1,900 to 4,200 years ago, followed by a shift to endogamy.

Narasimhan and colleagues refined this in 2019 with ancient DNA: the main ancestral population of modern South Asians formed from people related to early Iranian farmers and hunter-gatherers and a deep indigenous ancestry they called AASI, later joined by steppe-related ancestry from the northwest. Bangladeshi Bengalis sit on this same structure. The modern borders date from 1947 and 1971; the genome is thousands of years older.

The East Asian-related component

The clearest published measurement comes from Karlsson and colleagues, who in 2013 genotyped 42 mother, father and child trios, 126 people in all, of Bengali ethnicity from Dhaka. They found no evidence of substantial internal structure, estimated 9.3% (plus or minus 2.6%) East Asian ancestry, and found that mixture between Indian and East Asian sources best explained the population, dating it to 52 (plus or minus 2) generations ago, around 500 AD by their estimate. That is one sample from one city, and families differ, but it is the most direct figure available.

Where that share came from is still being worked out. Basu, Sarkar-Roy and Majumder reported in 2016 that mainland Indian populations carry four ancestral components, not two: ANI, ASI, an Ancestral Austro-Asiatic component and an Ancestral Tibeto-Burman component. Bengal lies between regions where the last two are common. Tätte and colleagues modelled Munda speakers, an Austroasiatic group of eastern India and Bangladesh, as about 29% Southeast Asian ancestry on average, with admixture dates of roughly 2,000 to 3,800 years ago. The later Dhaka date suggests at least some of the Bengali share arrived afterwards. The eastern component is real, modest in size, and probably has more than one source.

Chittagong Hill Tracts communities

Some communities in Bangladesh differ substantially from the Bengali majority. Gazi and colleagues studied 108 Chakma, 97 Marma and 97 Tripura individuals from the Chittagong Hill Tracts. All three showed strong affinities to Tibeto-Burman groups of Northeast India, while carrying substantially more mainland Indian ancestry than those groups. More than 60% of the men typed in all three communities carried the Southeast Asian-specific Y lineages O2 and O3. If your family is from one of these communities, a much larger East Asian share in a consumer result is expected and not an error.

Sylheti, Chittagonian and other regional backgrounds

Many British Bangladeshi families, especially in east London, trace their roots to Sylhet, so this is the question most readers ask. The honest answer: we could not find a published genome-wide study that samples Sylheti, Chittagonian, Noakhali or Barisal families as named groups. The main public references are the Dhaka sample above and the 1000 Genomes Bengali in Bangladesh panel.

Large datasets on British Bangladeshis do exist. The East London Genes & Health cohort profile reports 31,508 volunteers with ethnicity recorded, 67% of them Bangladeshi or British Bangladeshi. But that study is built for health research, and its published profile does not break participants down by district. So no service can honestly report a Sylheti percentage today. A careful analysis places you against the nearest sampled reference groups and tells you which you most resemble.

Haplogroups in Bangladeshi families

Your Y chromosome traces one paternal line and your mitochondrial DNA one maternal line. Neither measures your overall ancestry.

• Paternal lines. Most Bangladeshi men carry the same paternal lineages found across north and east India, such as R1a, H, J2 and R2. A minority carry East and Southeast Asian lineages such as O-M95, often written as O2a in older naming, which in South Asia is closely associated with Austroasiatic-speaking communities.
• Maternal lines. Most South Asians carry maternal lineages of deep South Asian origin, chiefly within haplogroup M and the R and U branches (see Indian maternal haplogroups). East Asian maternal lineages also occur in the region; in the Chittagong Hill Tracts study, haplogroup F1 was found in 12% to 24% of each community's samples.

A man with an O-M95 paternal line can have an overwhelmingly South Asian genome, so treat haplogroups as one thread, not a verdict. Whether yours can be read depends on your file containing Y and mitochondrial rows, which varies by company and chip version.

What your 23andMe or Ancestry result can and cannot tell you

What it can tell you

• That your ancestry is South Asian with the eastern share typical of Bengal, which is what the main label is really detecting.
• Who your close relatives are, if they tested with the same company.
• A basic haplogroup, if the company reports one.
• Most importantly, a raw data file: the measurement itself, roughly 600,000 to 900,000 genotypes, which can be re-analysed elsewhere.

What it cannot tell you

• Your district or village. No current reference panel separates Sylhet from Comilla or Barisal by DNA alone.
• Religion, caste or community. DNA does not record faith, and it cannot tell you whether a line was Hindu, Muslim, Buddhist or Christian at any point in history.
• Family traditions of Arab, Persian or Turkic descent. Every South Asian already carries Iranian farmer-related ancestry thousands of years old, so a small, more recent West Asian contribution is hard to separate from it on a consumer array. Trace Middle Eastern or Central Asian percentages are the least reliable part of any estimate.
• Exact proportions for groups nobody has sampled. A test can only compare you with the groups in its panel.

What changes when you upload the file to Helixline

The measurement is the same; the comparison changes. A global panel covers the whole world, so South Asia gets a handful of categories. Helixline compares your genotypes against South Asian reference populations, including 140+ community clusters, so the answer can be more specific. Our comparison of upload sites for South Asian ancestry sets out how the main options differ.

For a Bangladeshi file, that means:

• Deep components shown separately: AASI, Iranian farmer-related and steppe-related ancestry, plus the East Asian-related share that drives the Bengal label elsewhere.
• Closest reference groups. The sampled South Asian populations your DNA is most similar to. This is a similarity ranking, not a ruling on caste or community; where a group has no published sample, you see its nearest sampled neighbours instead.
• Y-DNA and mtDNA haplogroups, where your file carries the rows needed to call them.

The Ancestry report costs ₹1,999 / $25, and the Complete report (ancestry and health) costs ₹3,999 / $50. Results normally arrive within 24 hours of payment and a valid file, and your data is stored encrypted in India. Nothing is shipped, so it works the same from London, New York, Toronto, Riyadh, Dubai or Melbourne. See also our diaspora DNA guide and, for British readers, our UK diaspora guide, which covers the same testing problems.

How to download your raw data and upload it

23andMe

Open Settings from your profile name, scroll to 23andMe Data and request the raw data (not the report PDFs); you are emailed when the .zip is ready. 23andMe changed ownership in 2025, so download it now (see what to do about 23andMe). Our 23andMe raw data format guide explains the file.

AncestryDNA

On your DNA page, click the settings icon, choose Download Raw DNA Data, confirm with your password and click the link in the email. Full steps are in our AncestryDNA raw data guide.

MyHeritage

Under the DNA menu choose Manage DNA kits, open your kit's actions menu and download the raw DNA data; MyHeritage emails a link when it is ready. See our MyHeritage raw data guide.

FamilyTreeDNA and Living DNA

Both offer a raw data download from your account. FamilyTreeDNA may offer autosomal and X data as separate files, so take the autosomal Family Finder file.

Then upload it

1. Keep the file exactly as downloaded. A .txt or .zip, unmodified. Never open it in a spreadsheet and save it. If in doubt, run our seven checks for a complete raw DNA file.
2. Go to the Helixline upload page and click through to app.helixline.in.
3. Choose your report, Ancestry or Complete, and pay inside the app.
4. Upload the file. Files from 23andMe (v3, v4, v5), AncestryDNA, MyHeritage, FamilyTreeDNA and Living DNA are accepted.
5. Read your report, normally within 24 hours of payment and a valid file.

Frequently Asked Questions

What does Bengali & Northeast Indian mean on 23andMe?

It is one of 23andMe's South Asian reference populations. In its July 2024 update the company listed it as built from Assamese and Bangladeshi reference samples, and described it as one of its more geographically broad clusters. A high percentage means your DNA resembles those samples; it does not identify a district, a community or a religion.

Why do Bangladeshis get a higher Bengali score than West Bengalis?

Mainly because the category was built partly from Bangladeshi samples, and because it leans on the East Asian-related share of ancestry that is a distinctive feature of Bangladeshi Bengalis. People who carry a little more of that share are pulled towards the label. It says nothing about who is more Bengali.

Do Bangladeshis have East Asian DNA?

Yes, as a minority share. A genome-wide study of 42 Bengali parent and child trios from Dhaka estimated about 9.3% East Asian ancestry on average and dated the mixing to roughly 52 generations ago, around 500 AD. The rest of the genome reflects the same South Asian ancestry found across the subcontinent, and individual families vary.

Can DNA tell me if my family is Sylheti?

Not directly. We could not find a published genome-wide study that samples Sylheti families as a named group, so no test can honestly report a Sylheti percentage. What an analysis can do is show which sampled reference groups your DNA is most similar to, and confirm close relatives through matching.

Can I upload my raw DNA file from the UK, US or the Gulf?

Yes. Uploading a file involves no shipping, so it works from anywhere. Helixline accepts unmodified .txt or .zip raw data from 23andMe (v3, v4 and v5), AncestryDNA, MyHeritage, FamilyTreeDNA and Living DNA. The Ancestry report is $25 (₹1,999) and results normally arrive within 24 hours of payment and a valid file.

Conclusion

A Bangladeshi result on 23andMe, AncestryDNA or MyHeritage is usually right as far as it goes: South Asian ancestry with the eastern share typical of Bengal. What it hides is everything underneath, from your deep ancestral components to your closest reference groups and your paternal and maternal lines. Sylheti and other regional groups have not yet been sampled as named populations, and an honest report says so. The file you already have is enough to get that report, without a new kit.

Read next: Bengali DNA ancestry across Bengal.