Why even small African matches matter!

On 14 July Ancestry announced that they will pull the plug on the smaller matches they have been reporting for years now. In particular distant matches which show a shared DNA amount within the range of 6cM – 8cM. For many people this may effectively wipe out more than half of their total matches. Furthermore for Afro-descendants this could mean the loss of many invaluable African matches! Overall I would estimate that on average in between 50-75% of all African matches might be erased by Ancestry’s update.1 In one big drastic operation… and with just a few weeks notice…

An alarming situation for anyone who is reliant on their DNA matches (incl. smaller ones) for their research. Fortunately there are some strategies to preserve your smaller matches. Especially by contacting them, grouping them, starring them or adding notes. For more details see this overview of various helpful blog articles:

An artisanal miner holds an uncut diamond in her hand in Kimberley

A South African woman is holding an uncut diamond in her hand (photo credits) African DNA matches can be as illuminating as diamonds. Even when small in size they can still be priceless! However carefull assessment is a MUST when dealing with potentially false matches because otherwise you’ll only end up chasing fool’s gold 😉

***

In this blog post I will focus on the importance of smaller African matches. Something which has been revealed to me in my ongoing research on African DNA matches since 2017.2 Even when indeed many of them will be false matches. Precious informational value may still often be obtained. But it seems that this importance has not really sunk in yet with Ancestry or perhaps they have other priorities… Either way Ancestry’s announcement has been brutally abrupt and will catch many people by surprise. 

I hold no illusions about a complete reconsideration of this update (I love to be proven wrong though!). However I do hope the main arguments listed below may at the very least persuade Ancestry to delay their intended update. Even more so I wish that they wisen up and instead will implement tailor-made solutions which take into full account the diversity of their customer needs! If you agree please forward this blog post to Ancestry! For example in this suggestion box. See also the last part of this blogpost.

  1. Added value of small African matches
  2. Dangers of small African matches
  3. Appeal for true committment

 

Why even small African matches matter!

  • Afro-descendants face exceptional challenges when Tracing African Roots. Given the lack of a paper trail for the most part they have to go by any clues given! Out of sheer necessity the net has to be cast wide. An openminded and creative research approach is required. Both including but also going beyond genetic genealogy. All the while avoiding source snobbery in order to maximize any potential informational value.
  • Ancestry offers the biggest pond of African matches by far, when compared with other DNA testing companies. In great part because of its lower matching threshold which starts at 6 cM. This has been a precious resource for many which cannot be found elsewhere. It will be a tremendous loss when this pool of African matches will no longer be available or shrunk to a mere shadow of its former size!
  • Smaller matches may serve to reinforce more solid ancestral clues provided by related bigger matches (for example sharing the same meta-ethnic background). When combined with associated regional admixture and insights from other fields possibly leading to fruitful complementarity. See for example my survey of African DNA matches reported for 30 Jamaicans.
  • Small matches will indeed often be false matches or genealogically irrelevant but they may still be historically relevant (given correct interpretation). Leading to more insight of one’s African heritage within a greater timeframe. In particular when appearing with increased frequency. See for example my survey of African DNA matches reported for 50 Cape Verdeans.

 

1) Added value of small African matches

benefits

Similar to diamonds small African matches are precious because it is so hard to find even a handful of them. Not all matches will turn out to be the real thing indeed. However beggars cannot be choosers! With carefull assessment & follow-up research you can still obtain polished gems from what used to be only uncut pieces of rock laying within seas of sand.  (photo credits)

***

The very low odds of tracing back to Africa for African Americans were highlighted in 2017 during an episode of Finding Your Roots by Henry Louis Gates Jr.. The guest of his show, musician and producer Questlove, was found to be descended from one of the last enslaved Africans to arrive in the US on a slave ship.

____________________

“The discovery of ancestors on the Clotilda isn’t just an interesting genealogical fact. As Gates says, it means that Questlove is the only African-American he knows who can answer a question that many have asked: not only where in Africa his ancestors came from, but how exactly they got to the U.S. in the first place.” (Time, 2017)

____________________

This certainly was a very remarkable finding! But such a verifiable paper trail leading all the way back to Africa must be extremely difficult to reproduce for ordinary African Americans. For one thing they will not be assisted by professional teams of historians and genealogists. Furthermore this finding only concerns one particular family line (relatively recent) among possibly hundreds of others. All individually to be traced back to several parts of Africa! How is a layman expected to ever uncover a majority of these lines, let alone one single one?  This (near) impossibility of the genealogical route for me underlines once more how there is a lot more at stake for African Americans and other Afro-descendants when taking a DNA test than for people with plentiful documented knowledge about their ancestral origins.

With nothing concrete to go by it should be obvious that small African matches represent very valuable leads for anyone wanting to Trace African Roots! Due to lack of viable alternative Afro-Diasporans cannot afford to be too picky about the imperfections of DNA matching. It may perhaps be true that small African matches will almost never be fully validated with all the scientific rigour required. And many of them will indeed prove to be false or random matches (IBS= Identical by State). And even when inherited from your parents (IBD= Identical by Descent) these small matches may still be traced back further in time than imagined. That is more than a thousand years ago rather than within a genealogically meaningful time frame of let’s say 500 years. Such matches are also sometimes termed population matches (IBP= Identical by Population). 

Either way there will always be valid matches among these small African matches as well. Even while representing a minor share they will still hold incredible value because of their rarity! Merely being able to reach out to any African DNA match at all is of immeasurable significance to many people already! In principle I would agree that a cautious research approach is called for because it is so difficult or even impossible to make the distinction with false matches. But instead of treating all small matches as poison to be avoided at all costs I think the more fitting analogy for Afro-descendants would be that we are searching for diamonds by sifting through the sand, when Tracing African Roots. The risks of possibly “contaminated” research outcomes to be preferred over no research outcomes at all. I will deal with the dangers of small African matches further below in section 2.

It has been a true blessing in the last couple of years that Ancestry’s huge customer database (18 million at latest count!) has been including an ever growing number of Africans to get matched with. Combined with the lower matching threshold (6 cM) the odds of finding African matches were getting better every year. From my observation the chances of finding African matches were still very low in 2017. Probably within a range of 0-10 with smaller matches (6-8cM) representing a share of atleast 50% or greater. 

However from my current findings I would say that African Americans will typically receive around 20-30 African DNA matches! Among thousands of other matches. More than half of these African matches will again tend to be below 8cM. These estimates are based on having analyzed the DNA matches for about a hundred African Americans by now, using my filtering & scanning method. See also this blogpost for detailed overviews of African DNA matches I found for 2 African Americans in 2017:

A very different situation arises when being tested by 23andme or MyHeritage from my experience. These DNA testing companies already apply a more stringent matching threshold (see this overview). And in the case of 23andme also apply a restrictive cap on the total number of DNA matches. This might be beneficial for those people who are only interested in their closest matches. Because they have plentiful other resources to fuel their research. However for Afro-descendants it is simply a MUST to also have access to their distant matches. Otherwise in many cases they will just not be able to perform any kind of research at all, when wanting to Trace African Roots.

I have only occasionally looked for African DNA matches on either 23andme or MyHeritage. Basing myself on about a dozen observations for Afro-Diasporans I would say the odds of finding native/unmixed African matches on those two platforms are very low. In fact often you will not not find any African match at all. The highest number I have personally seen only being 2 African matches. All the more adding to the great value of Ancestry’s pool of African DNA matches. Up till now that is. See also:

 

Reinforcing other more solid ancestral clues

Table 1 (click here to enlarge)

WA matches

This table is showing the West African subset of all the 861 African DNA matches I managed to find for 30 Jamaicans tested by Ancestry. Nigerian & Ghanaian predominance is clearly reflected in their DNA matches. This goes both for the smallest matches (<7cM) as well as the matches greater than 10cM. The overall matching patterns serving to corroborate other ancestral clues such as regional admixture, slave trade history,  cultural retention and maternal haplogroups.

***

This section and the following one are not meant to be exhaustive. But rather to illustrate the main arguments I have listed in the beginning of this blog post. Based on my own research on African DNA matches, performed since 2017.3 To start with my most recent survey of African DNA matches being reported by Ancestry for 30 Jamaicans. Among many other findings from that survey the most significant outcome probably was a predominance of Nigerian & Ghanaian DNA matches in line with prior expectations. Both among the smaller matches (<7cM: 172 + 58) as well as the more solid matches (>10 cM: 82 + 37).4 Serving to corroborate other ancestral clues hinting towards Nigeria and Ghana being the most important places of African origins for Jamaicans. Such indications derived from regional admixture, slave trade history,  cultural retention and maternal haplogroups. For more details see also:

For the group as a whole I suppose the bigger matches (> 10cM) would already suffice to make this point. However on an individual level it was not always the case that my survey participants received either Ghanaian or Nigerian matches greater than 8 cM! The mere fact that there should be so many small matches from Ghana and Nigeria as well is surely no coincidence. As this would be in line with expected dilution from ancestors to be traced back mostly to the 1700’s (see this chart taken from the Slave Voyages Database).

Even when false matches may very well be among them the geographical distribution of these smaller matches was not randomized across Africa at all. But rather focused on the so-called Lower Guinea area within West Africa. In particular Ghana and Nigeria. Do also notice from Table 1 that so-called close matches (>20 cM) were very rare. Only 5 out of a total of 861 matches, which is a share of less than 1%. In order to reach a higher level of robustness really the entire dataset, incl. smaller matches, was indispensable for my research purposes. Combined findings being indicative of overall matching strength.5

Let alone for the average DNA tester for whom it may be much more difficult to obtain a complete overview of African DNA matches on Ancestry. And therefore any match from Ghana and Nigeria, either small or big, would be very valuable. The smaller matches may not always have been genealogically relevant indeed. I suspect many of them could be population matches especially. But even so they still point in the right direction of identifying the modernday country associated with one or more family lines. Backed up by historical plausibility, regional admixture and other ancestral clues pertaining to Jamaicans as a group.

Table 2 (click here to enlarge)

DNA matches1

This overview  shows the Nigerian DNA matches I found for 1 single Jamaican. It is highly illustrative of how your Nigerian DNA matching patterns may enable further ethnic specification of your Nigerian lineage. Take notice how the abundance of most likely Igbo matches (going by surname) is reflected among both bigger and smaller matches. I have left out name details for privacy reasons. While the regional summaries are reflecting Ancestry’s flawed update in September 2018. Currently all these matches will show a much more convincing “Nigeria” score. Often showing up as only region (see this blogpost).

***

The frequency of smaller matches (< 7cM) also proved useful when drilling down to plausible ethnic background of these African DNA matches. As the smaller matches usually were in strong correlation with the bigger matches. Adding to overall coherence. Actually my research goal was not to find a “certified” basis for ethnic identification. Something which I believe is wrought with all sorts of complexities.6 Instead I was attempting to get a general idea by processing the data on a best effort basis. And to atleast make plausible inferences on a broader meta-ethnic level.

Keeping in mind the disclaimers I discuss in greater detail in my original blogpost it was still very striking to see a high number of especially Nigerian Igbo matches. Most likely due to substantial ancestral connections with the Bight of Biafra. As supported by plentiful historical evidence for Jamaica. Then again most people did have Nigerian DNA matches from other ethnic groups as well. In particular Yoruba ones. Some of these matches were quite small though (<7 cM). And therefore presenting greater chances of being mere population matches (IBP) or even just false positives. Still in most cases they should be indicative of generic roots hailing from southern Nigeria.

The intriguing thing was that during my survey I was actually able to establish increased odds for three of my individual survey participants having substantial Yoruba lineage in addition to more prevailing Igbo lineage. First of all by taking into account their entire matching patterns (incl. smaller matches) and by comparing what was typical for the survey group as a whole. But also then looking into other clues provided by regional admixture, parish origins etc. Naturally more follow-up research would be required to build a more solid case. But still these smaller matches provided very promising leads which are hard to come by and therefore not to be discarded! For more details:

 

Illuminating wider historical context

Table 3 (click here to enlarge)

NorthAfrican matches

This table is showing the North African subset of the 437 African matches I managed to find for 50 Cape Verdeans tested by Ancestry. Notice that the greater part of these 180 North African matches (64%) is below 7cM. It may reasonably be expected that a large part of these small matches are either IBS or population matches (IBP). Therefore not genealogically relevant. However from a historical perspective this increased frequency of North African matches does make sense. And also serves to corroborate minor but distinctive North African admixture scores reported for Cape Verdeans.

***

As already noted small matches will indeed often be false matches or genealogically irrelevant (within a timeframe of around 500 years). However especially so-called population matches (IBP) may still be historically relevant. Leading to more insight of one’s African heritage within a greater timeframe. In particular when appearing with increased frequency. However this does require very careful interpretation and a solid basis of historical knowledge. Mere wishful thinking will lead to random speculation and self-deceipt!

In my survey of African DNA matches being reported by Ancestry for 50 Cape Verdeans one very remarkable outcome was the increased frequency of North African matches. North African matches were reported for 80% (40/50) of my survey participants. Also the total number (180) is surprisingly high. Representing a share of 41% (180/437) of the total number of African DNA matches. While in my Jamaican survey only 3 North African matches turned up out of a total of 861 matches. Representing a minimal share of less than 1% (3/861=0.3%, see this overview). So clearly this outcome was not just some fluke but quite consistent already. However the average size of these matches was minimal at around 7cM. A majority of these matches (63%) actually being even smaller than 7cM!  

Correct interpretation of these matching patterns is crucial. First of all it needs to be realized that this relatively high number of North African matches is clearly in contradiction with actual ancestral proportions. Reviewing the associated regional admixture levels for my survey group (on average 46% for “Senegal” +”Mali” versus 0.5% for “Northern Africa”) is particularly useful in this regard to get a proper sense of proportions. On the other hand the occurrence of these North African matches is certainly not unexpected for Cape Verde! As historical evidence does back up actual North African ancestry for Cape Verdeans. Take note also in Table 3 that the geographical distribution within North Africa is clearly gravitating towards the western Maghreb and is not randomly spread towards Egypt.

Again many if not most of these smaller matches may very well not be traced back within a genealogical timeframe of 500 years. However I suppose for population matches mutual ancestry may be indicated going back further in time. Let’s say 1000 years when most of Portugal was still under Moorish occupation. And also the Fula ancestors of Cape Verdeans might have intermingled with Berber North Africans to some degree around that time. Of course in such instances it will be fruitless to actually identify any shared ancestor. However such a finding does provide an extra historical perspective on your background. Also it serves to corroborate minor but distinctive North African admixture scores reported for Cape Verdeans. An aspect of their genetics which they share with many Latin Americans (see this chart).

In fact for Cape Verdeans several ancestral scenario’s apply when dealing with North African matches. Including relatively recent ancestry from North Africa. As indicated also by the biggest match within my survey being 17.9 cM (see table 3). Reported for a person with a known Moroccan Jewish ancestor from the late 1800’s. Another person received no less than 16 North African matches. However all of these matches were no greater than 8 cM (see CV13 in this sheet). And therefore I suspect to be traced back further in time. False matches might very well be among them.

Actually if we review table 3 again it can be seen that the overall share of North African matches greater than 8cM is only 13% for my entire survey group. Implying that nearly 90% of these informative matches will no longer be visible after Ancestry’s intended update. My survey participants would surely agree though that these mostly smaller North African matches represent an insightful outcome you would not want to miss out on. Even when one must be aware that a wider historical timeframe will apply. For more discussion see:

 

2) Dangers of small African matches

____________________

“ It is probably no exaggeration to say that for many Afro-Diasporans getting connected with an African DNA cousin represents nothing less than a dream come true. It is considered a highly prized outcome. As a genealogical research reward in itself but often also on an emotional or even spiritual level. Understandably this sometimes leads to wishful thinking and tunnel vision whereby DNA results are not critically assessed.” (Fonte Felipe, 2017)

____________________

The above quote is taken from my tutorial about a scanning & filtering method to single out African DNA matches (see this link). Posted originally in 2017 and I am still 100% behind this statement. Again without wanting to rain on anyone’s parade I do feel that the awareness that smaller African matches will also include false matches is often lacking. For several understandable reasons. However I do therefore consider it my moral responsability to keep pointing out these shortcomings of DNA matching. Which I have attempted to do to the best of my knowledge whenever blogging about my ongoing research on African DNA matches since 2017. Similar to how I have also tried to always caution against the limitations of admixture analysis in my previous research efforts (see this overview).

I suppose two opposing views may exist in regards to African matches. Either one believes all of them (incl. small matches) are the real deal and therefore IBD (Identical By Descent). To be fully embraced as family from the moment of discovery. But others may be more cautious and assume smaller African matches are possibly false positives or population matches until proven otherwise. These are rather extreme positions of course. But it seems to me that especially in the absence of a historically plausible context a more conservative approach is called for. Understandably the eagerness for African matches among Afro-Diasporans is very high. But I feel it is wise to resist the urge of finding a fitting ancestral scenario right away whenever a more surprising African match is reported. As this could be a form of self-serving confirmation bias.

I suppose it is also a matter of personal preference whether to focus on historically plausible matches in line with the majority of your regional African admixture. Or rather to go on a wild goose chase of African matches from more eccentric and unexpected places. Especially when unsupported by any other clues. Some people tend to have a particular infatuation for East African matches. Often just simply out of natural curiosity because such findings are so surprising. But at times also because of ideological reasons. However direct ancestral ties between the Atlantic Afro-Diaspora and especially Northeast Africa seem least likely to be supported by any documented evidence I am aware of. Granted: absence of evidence is not evidence of absence. Then again historical plausibility does greatly impact probability.

Because of Ancestry’s current threshold it does occasionally happen that African Americans and other Atlantic Afro-descendants receive small matches from especially countries like Kenya, Uganda or Sudan and more rarely even from Ethiopia. When these matches are taken at face value people then run the risk of “claiming” a certain type of exotic lineage which is neither genealogically nor even per se genetically valid. Something which is bound to happen when you are not taking into account the impact of both ancient and historical migrations across the African continent. The Bantu migrations from Nigeria/Cameroon into Central, Southern and Eastern Africa being a great example of how ancient migrations may lead to population matches all over the place. As well as the more recent Fula migrations from Senegambia all along the Sahel corridor into Sudan. Basing myself on my own research these kind of connections can often explain the occurrence of unexpected smaller matches from East Africa.

To keep things in perspective though it seems that these historically unexpected matches only make up a very small part (<5%) of the total. Based on my research sofar an overwhelming majority of African small matches will still be in line with historical plausibility. And therefore not totally random! Even when many of them will be genealogically irrelevant within a timeframe of 500 years.

In my Jamaican survey the number of unexpected North & Eastern African matches was very uncommon. Only around 1% (12/861) of the total. And in fact some of these matches were still useful because I was able to make a plausible connection with Fula migrations into Sudan. Also in my Cape Verdean survey the number of unexpected matches from outside of West Africa was minimal: 15/437= 3%.  And again some of these smaller matches (from Central & Southern Africa) were still informative as I was able to find additional clues which made a possible detour by Brazilian lineage likely. For more discussion see:

 

Mozambican ancestry may lead to matches from both Southern & East Africa

Map 1 (click here to enlarge)

Map STP

This map shows the African matches on Ancestry for my DNA cousin who is half Mozambican. His partial Cape Verdean connection (by way of São Tomé) resulted in the greatest number (539) of African DNA matches by far. Otherwise most of his African matches are from Southeastern Africa. However due to a skewed customer database as well as ancient/modern migrations the most numerous mainland matches were from South Africa, Zimbabwe and Kenya. And not from Mozambique. This outcome could have been misleading if this person was not already aware of his actual background. 

 ***

Last year I was kindly given the opportunity to do a full scan and analysis of the DNA matches on Ancestry for my DNA cousin whose father is from Mozambique, in Southeast Africa. Actually we are related through his mother who is from São Tomé & Principe. An islandstate in West Africa (see map above). Because she happens to have either a Cape Verdean parent or grandparent. So similar to people from the Afro-Diaspora he has African ancestry from several corners of the continent! Which is also clearly reflected in his African DNA matches shown above. However had he not known in advance about his actual background his African matches could have been misleading. Both the bigger and smaller ones actually. Afterall his recent African ancestry hails from only three countries. But his African matches hail from no less than 19 countries!

Especially his Mozambican connection is not prominently showing up in matching strength. His most numerous group of African matches by far being Cape Verdeans (539). Even when he probably is only 1/4 or 1/8 Cape Verdean. Otherwise the greatest number of mainland African matches was from South Africa, Zimbabwe and Kenya (in that order). This outcome is to be explained first of all because of a skewed customer database effect on Ancestry. Whereby the number of Cape Verdean-American customers  as well as African migrants from former British colonies clearly is much larger than the number of DNA tested Mozambicans (as well as people from São Tomé & Principe). Which is simply due to their migrant presence in the USA and other parts of the West.

Then again both ancient and relatively recent migrations across Southern & East Africa do also account for his matching patterns. As well as the Indian Ocean Slave Trade. After close scrutiny and consulting relevant history resources most of his matches do make sense therefore. Even the smaller ones, which are usually from surrounding countries. However in particular the small matches from Uganda, Ruanda and Kenya are leading astray from his confirmed Mozambican heritage it seems. Especially when going only by intuition.

And again this outcome could have been even more so misleading if he had not been aware of his actual African origins. Which is the unfortunate situation for practically all people in the Afro-Diaspora. Taking such small DNA matches at face value would then indeed imply you are starting from a false premise and therefore likely to be ending up with a false answer. For more details see:

 

Small African matches: potential diamonds or poisoned M&M’s?

dangers

African DNA matches, incl. smaller ones, may be seen as diamonds in the rough which need to be sifted from seas of sand & rocks with great scrutiny. And afterwards also to be polished by rigorous follow-up research. In order to avoid self-deception it is a MUST to be aware of the dangers of false matches throughout this process. Photo credits: CNN

____________________

“I consider small segments to be “poison,” in that too many of them are false matches and we can’t tell the difference between the false segments and the real segments. I use “Poison M&Ms” as an example. If I handed someone a bowl of M&Ms and told them that 30% are poisoned and there’s no visual difference (similar to 30% of small segments of 5 cM or smaller), no one would eat the M&Ms. Similarly, we can’t use small segments because they poison our genealogical conclusions.” (Genetic Genealogist, 2017)

“Indeed, it will be important to consider and study how this might affect people with African or other historically marginalized ancestry. All the available science about small segments indicates that this change will actually improve the method by weeding out false data and preventing incorrect conclusions.(Genetic Genealogist, 2020)

“Additionally, as more Africans test the number of larger shared segments will increase. Those of us that have been genetic genealogists for a long time have seen several instances of DNA results provided to people of African descent that were later discovered to be incorrect as testing improved and databases grew. Hopefully this change prevents that from happening.”  (Genetic Genealogist, 2020)

____________________

The analogy of small DNA matches being like a bowl of poisoned M&M’s has been coined by the Genetic Genealogist. I would like to underline that I am largely in agreement with his concerns highlighted above. In particular the need for continued awareness of false matches. As well as resisting the temptation to jump to conclusions. See also these useful blogposts from which I myself have learnt a lot as well:

However all analogies are of course simplifications of more complex and also diverse realities 😉 Therefore I have come up with an additional analogy. In my opinion moreso corresponding with the particular situation of Afro-descendants in their challenging quest to Trace African Roots. The process of finding African DNA matches, incl. smaller ones, to be seen as diamonds in the rough which need to be sifted from seas of sand with great scrutiny. And afterwards also to be polished by rigorous follow-up research.

All the while being aware of the dangers of false matches. But not being completely dismissive either because despite imperfections even false matches (especially population matches) may still contain informational value as well. Plus there will be valid matches too. Therefore judging each case on its own merits, whenever possible. Of course this is again an imperfect analogy but from my experience more in line with the perspective of many Afro-descendants on this topic. 

Actually I do not think this is a binary question at all. Rather I believe these analogies represent different research agendas as well as different research approaches. Speaking for myself I have always said that basically I am a guy who likes to see the glass as half full rather half empty. Avoiding source snobbery and maximizing informational value despite shortcomings (which do need to be properly accounted for). Basically taking any promising lead I can get and combining with other clues from different fields. Not per se aiming for conclusive statements but rather plausible inferences.8 

In short my research agenda both includes but also goes beyond genetic genealogy. Across the years I have performed several surveys in order to find out to what degree DNA testing for Afro-descendants (both regional admixture and DNA matches) may correspond with expectations based on historical plausibility (especially slave trade patterns).9 In this sense my research angle may be said to have greater affinity with population genetics as it is mostly based on aggregate data.

However I do strongly believe that the implications of my research are also useful for individual DNA testers, when Tracing African Roots. Afterall as soon as your research as an Afro-descendant takes you across the Atlantic Ocean it will be a different ballgame! Because it will no longer be about 100% validation of DNA matches in order to complete extensive family trees, free of any possible errors. Instead it will be about finding light in the dark and restoring ancestral ties which were brutally disrupted due to Slavery. Attempting to piece together a tangible picture of your multi-layered African heritage. 

This quest can however be problematic when you have unrealistic expectations about identifying the actual names, birth villages or the exact ethnic identities of African ancestors (see footnote 6). As discussed above small African matches can be dangerous especially when taken at face value and also in isolation of other types of ancestral clues. However from my observation truly misleading and seemingly randomized DNA matches from historically unlikely places (such as East Africa) only represent a very minor share of the total number of African matches you may expect to receive on Ancestry.

And even then informational value might in some cases still be obtained with careful analysis. I therefore believe that when using sound judgement these risks of misconstruing smaller African DNA matches are limited and acceptable. I do think more educational context should be provided by DNA testing companies. Perhaps smaller matches should also be visually presented in such a manner that people will immediately be aware of what they’re dealing with. Although of course everyone also has their own responsability to inform themselves properly.

Given that most Afro-descendants trace back their African ancestry to the 1700’s or even earlier it should not be surprising that many of their African DNA matches will tend to be smaller than 8 cM. While the odds of finding close matches (>20 cM) will usually be very slim. As obviously across the generations the amount of shared DNA with distant living relatives in Africa has greatly diminished. Again there will be many false matches among them however this 6cM-8cM range also simply happens to be the range where many of their valid DNA matches are to be expected! See also:

Instead of a complete dismissal of smaller matches my approach would be to first of all analyze your main African matching patterns. Basing yourself on a complete overview of all your African DNA matches. Carefully looking into relative frequencies and clusters of matches with a certain background which might also align with bigger matches. In order to single out the more robust findings. Rather than allowing your research to be distracted by “exotic” outliers among your matches.  

Next step would be to combine your entire overview of  African DNA matches (incl. smaller matches) with regional admixture (within Africa). From my experience this will often lead to enhanced insights and complementarity. Add in proper historical & genealogical research and your multi-faceted strategy to Trace African Roots will be optimized. Which will allow you to obtain more clarity about your African origins. Especially when it comes to these aspects:

  • Corroborate (macro)-regional admixture 
    • In my African DNA matches surveys sofar I established a strong correlation in particular with so-called macro-regional admixture. For Cape Verdeans I found that the Upper Guinea area was the most likely origin of around 88% of all African matches (south of the Sahara). Which correponds nicely with their combined “Senegal” and “Mali” scores being around 92% (within their scaled African breakdown). While for Jamaicans I found an 87% proportion of Lower Guinean related matches (mostly Ghana & Nigeria) out of all African matches. Which is roughly in line with the almost 70% Lower Guinean share I found during a previous survey based on regional admixture for Jamaicans. 
    • In upcoming surveys I will investigate if a similar historically justified predominance for Central Africa in both DNA matches and regional admixture will be obtained for Brazilians and Haitians. Also for African Americans this can be a fruitful approach. Although they tend to be more evenly balanced in their (macro) regional origins within Africa. Naturally several limitations are to be kept in mind. For an overview of disclaimers see this page.  
  • Zoom into plausible (meta-)ethnic background of specific family lines 
    • During my African DNA matches surveys sofar it has been quite striking how certain (meta)-ethnic backgrounds were very commonly reported. For Cape Verdeans this happened to be especially Fula/Fulani matches from the wider Upper Guinea area. While for Jamaicans it was firstmost Igbo matches from Nigeria. Additionally also the distinction between possibly Akan and Ewe matches from Ghana was made possible by taking into account their surnames (as wel as regional admixture, see methodology section in this post). These kind of matches are also regularly reported for African Americans from my observation. And in addition an increased number of Malagasy matches is often also notable for African Americans (see footnote 3).
    • Whenever matches with the same plausible background are being reported with increased frequency it might be very helpful to perform dedicated follow-up research. Especially when shared segment info is available from other DNA testing platforms you might then apply triangulation as well as DNA Painter with great effect. Comparing family trees with other DNA matches who also share the same cluster of African matches with you might then enable you to identify an earliest family line associated with these matches! 
    • Such an effort offers great potential in my opinion. Naturally to be combined with any other clues you might have. Also it goes without saying that extra scrutiny is required! Then again even false matches (in particular population matches shared with close family) may still contain informational value and offer a chance to advance your research. Hopefully leading to an end of total obscurity about your African ancestors. At least along one family line. The end result may admittedly often not be 100% conclusive but will still have been corroborated to the greatest extent. Shining through like a polished diamond! 
  •  

3) Appeal for true commitment 

 

real or fake

This statement was made by Ancestry’s CEO on June 3, 2020. It is good to know that the intentions for making much needed improvements to Trace African Roots seem to be in place already. However as always actions speak louder than words…See this link to Ancestry’s blog. As well as this link to Ancestry’s Facebook page (where you can also leave a comment).

 

____________________

“I am of Cape Verdean descent myself and I am personally not expecting any special treatment or favours from profit-driven companies such as 23andme or Ancestry. Even when I can easily imagine that as an USA-based company some degree of consideration for African Americans is in order. […]

Most people in the Afro-Diaspora do not have any detailed knowledge about their African roots and are usually very eager to learn more. Not anything with “100% accuracy” even but just something meaningful and relevant which goes beyond the lump category of “West African”. […]

“Unfortunately it seems that many DNA testing companies are either not able or do not have a true commitment to cater to the particular needs of Afro-Diasporans when it comes to admixture analysis and other aspects of DNA testing. I always thought Ancestry was an exception but I might have to change my mind after this upcoming update… ” (Fonte Felipe, 2018)

____________________

In a recent statement triggered by the Black Lives Matter protests Ancestry said that they are committed to “make our product experience as inclusive as it can possibly be for everyone“. Now more than ever is the time to show these are not just empty words! I would urge Ancestry to stop the intended deletion of small matches. And to continue the reporting of small matches which also can be clarifying within the 6 cM-8cM range. As this blogpost has demonstrated that even small African matches matter! Based on my research the added value of these matches is much greater than the associated dangers of false matches. Like myself many other Afro-descended researchers have only just got started in exploring these precious ancestral clues to the fullest! As far as I am concerned the argument that this update will be beneficial for all of Ancestry’s customers clearly does not hold!

Although not confirmed it has been widely speculated that Ancestry has ulterior motives for this update. In particular the reduction of operational costs of its main website where your DNA matches can be consulted. Which is understandable in a way because of Ancestry’s huge customer database. However if for some reason this update is deemed unavoidable then Ancestry should atleast give people sufficient time to deal with these changes. Preferably till the end of the year but I would recommend at least a couple of months. As many people only log in irregularly into their Ancestry account. In fact many may now be on a summer holiday!

As mentioned in the beginning of this blog post you can preserve your small matches by contacting them, grouping them, starring them or adding notes. This can however be a laborious process! For more details see this overview of various helpful blog articles:

If this update is indeed inevitable then Ancestry should also seek to compensate for this loss by offering new tools geared to facilitate specialized research for Afro-descended customers. In 2018 I already made an appeal for true committment on the occasion of Ancestry’s update of their Ethnicity Estimates in 2018. Which sadly turned out to be a major setback. Because Ancestry’s pioneering African breakdown lost most of its informational value then (see this blogpost). The following update in 2019 only partially managed to recover some of the damage (see this blogpost). However essentially still no substantial progress has been made for Atlantic Afro-Diasporans. 

The suggestions for improvement I made in 2018 have largely been ignored by Ancestry up till now. If Ancestry’s stated intentions are truly serious this time they should start listening and acting on it! Across the years many people have urged Ancestry for a whole range of useful improvements. To back up their own words Ancestry should implement atleast one of these following suggestions. But preferably of course progress should be made on several fronts! To ensure that Ancestry’s experience from now on will be truly inclusive!  

  1. Preserve current small DNA matches which have been reported up till now for those people who wish to keep them. These were afterall part and parcel of  Ancestry’s original DNA testing package people paid good money for!  
  2. Allow optional settings for continued reporting of small DNA matches within the 6-8 cM range 
  3. Create Genetic Communities for West & Central Africa. In particular for Nigeria, Ghana and Liberia, as sufficient customer samples may already exist for those countries. And also for Cape Verde as an independent country!
  4. Enable DNA matching with samples from Ancestry’s Reference Panel. Possibly then even a broader Upper Guinean and Central African genetic community could be realized?
  5. Provide a Chromosome Browser (sign this petition).
  6. Show Shared Matches at a lower threshold than 20cM.
  7. Show triangulated DNA segments
  8. Show ethnicity/admixture of shared DNA segments with your matches.
  9. Provide more ways of advanced filtering of DNA matches (incl. on ethnicity and genetic community)
  10. Implement other suggestions listed in my previous blogpost from 2018

 

Ivorian man connecting with his African American DNA cousins

This is a REALLY interesting short documentary by an Ivorian film maker who got tested by Ancestry and was reconnected with several African Americans. His background is Bété, which is a Kru subgroup (also found in Liberia). The amazing thing is he received about a dozen 4th cousin matches with shared DNA being >20cM. These matches are most likely to be traced back to relatively late shared ancestors from the 1800’s.

However in addition he also has hundreds of distant DNA matches, mostly African Americans! Going by historical plausibility and whenever valid these matches are most likely tracing back to the 1700’s because of greater dilution. Overall less reliable than the close matches but still incredibly precious! Even the population matches among them might still be useful when combined with additional clues. In particular when wanting to corroborate more generic roots from Liberia/Ivory Coast. Naturally one does need to refrain from jumping to conclusions 😉

It is truly a SHAME that most of these small matches will disappear within a month!!! When in fact this man tested exactly with the intent to reconnect with his matches! He is even actively encouraging other Africans to do DNA testing! Please contact Ancestry and let them know you want them to stop the deletion of small matches!!!

***

If you are not happy with Ancestry’s current plans let them know about it!!! Also forward them this blog post (when in agreement of course). Naturally you are free to personalize your feedback and add your own suggestions as well! In fact I would also very much like to encourage my blog readers to discuss these suggestions in the comment section below. In order to get a fruitful and constructive exchange of ideas going. If Ancestry truly values each demographic segment of their customer database they should take notice this time! 

 

Inspiring blog posts on African DNA matches

 

Africa diamonds

***

 

___________________________________________________________________________

Notes

1) This estimated loss of total matches (56%) is based on the share of DNA matches within the 6-8 cM range currently being reported by Ancestry for 24 persons. The individual range going from 44% to 72%. See also:

This corresponds with my own findings based on the African DNA matches surveys I have performed for 50 Cape Verdeans in 2018 as well as for 30 Jamaicans in 2019. In addition it also aligns well with what I have observed for about 100 African Americans whose African DNA matches I have analyzed since 2017 using my scanning & filtering method. Overall I would estimate that in between 50-75% of all African matches might be wiped out by Ancestry’s update.

2)  In May 2017 I blogged about a scanning & filtering method to single out African DNA matches. Since then I have been shifting my attention towards DNA matches and also how these may correlate with regional admixture. See also: 

3) Another powerful illustration of how small African matches can be reinforcing other ancestral clues would be based on the appearance of Malagasy matches for African Americans. These research findings are however more preliminary on my part. At least for now. I do intend to publish a detailed blog post about my findings eventually. In short these would be:

  • Malagasy matches from Madagascar are being reported for African Americans  with increased frequency when compared with other parts of the Afro-Diaspora
  • African Americans with Malagasy matches often tend to have (ultimate) Virginia state backgrounds
  • Malagasy matches tend to be small (<8 cM) and intriguingly are often themselves also showing up in Ancestry’s “Early Virginia African Americans” community
  • All of which corresponds very well with a historically plausible scenario of slave trade between Madagascar and the USA focusing mosty on Virginia and taking place especially in the early 1700’s
  • Founding effects and subsequent dispersal due to Domestic Slave Trade causing Madagascar matches to also turn up for people from the Deep South.
  • Minor but distinctive Southeast Asian admixture scores (more reliably shown on 23andme) also fall in line with expected dilution dating from the early 1700’s
  • Many people who receive Malagasy matches also receive small matches from other parts of the Malagasy Diaspora. Such as Mauritius, Réunion, Seychelles, Comoros, South African Coloureds and even Saint Helena. Because of their mixed background (incl. European admixture) also other ancestral scenarios might be possible though.
  • Follow-up research based on identifying the ethnicity of the actual shared DNA segments and triangulation could clarify things further

For a pathbreaking blog series on this topic see also:

The author of this blogseries, Teresa Vega, is currently writing a highly relevant book on this topic:

  • The DNA Trail from Madagascar to the Americas (to be published soon)

Overview below is taken from my African DNA Matches Service for an African American. It may serve as a further illustration of insightful matching patterns when both small and big matches are taken into consideration (also for the Fula/Fulani matches!). Take notice only one Madagascar match is >8cM. After Ancestry’s update the other most likely related matches (marked in red) will be lost! Even when several of them are also showing up for the son of this person (IBD). The small exotic matches from Northeast Africa and Kenya are indeed likely to have been false matches and also misleading. Except possibly the IBD one, but I suspect it is still an (ancient) population match either way.  However the small matches from Madagascar and Mauritus are historically plausible and quite likely genealogically relevant. With adequate follow-up research they could potentially help to zoom into shared DNA segments and even identifying associated family lines!

***(click here to enlarge)

DNA matches

4) In my Jamaican survey I applied a threshold of 7cM instead of 8cM to single out smaller matches. The share of total matches within the range of 6-8cM was 59% (= 510/861). See also this spreadsheet which contains all 861 African DNA matches I found:

5)  This was something I already came across when 23andme’s Country’s of Ancestry tool was still in place. Naturally one had to make allowance for a skewed customer data base (some countries being better represented than others) as well as other limitations. Still I found it very useful to see how often the top ranking order for the smallest DNA matches was in line with historical plausibility, as well as known recent family origins. And also usually following the ranking order obtained for bigger matches (set at 10 cM or even 15cM). Combined being indicative of overall matching strength. More randomized countries also being shown among the smaller matches (especially when set at 5 cM). Puzzling and misleading when taken at face value. But usually with decreased frequency and not among the top countries being reported. See also:

6) Understandably many people desire to have the most specific degree of resolution when searching for their African roots. They want to be able to pinpoint their exact ethnic origins and preferably also know the exact location of their ancestral village. In a way following in the footsteps of the still very influential ROOTS author Alex Haley. Unfortunately these are rather unrealistic expectations to have in regards to DNA testing. Not only given current scientific possibilities. But also because such expectations rest on widely spread misconceptions about ethnicity, genetics, genealogy as well as Afro-Diasporan history.

Too often people ignore how the melting pot concept is really nothing new but has always existed! Also in Africa where inter-ethnic mixing has usually been frequent! Throughout (pre) history and maybe even more so in the last 50 years or so. Generally speaking ethnicity is a fluid concept which is constantly being redefined across time and place.

Too often people fail to take into consideration how due to genetic recombination our DNA will never be a perfect reflection of our family tree but might actually also at times suggest very ancient migrations.

Too often people underestimate the actual number of relocated African-born ancestors they might have (dozens or even hundreds!). As well as the inevitable ethnic blending which must have taken place across the generations.

Too often people are still not informing themselves properly about Africa itself and the documented origins of the Afro-Diaspora. Many specific details may have been lost forever but there is a wealth of solid and unbiased sources available which can help you see both the greater picture as well as zoom in more closely to your own relevant context. See also:

7) The frequency of population matches, IBP is more common than many people might expect. Many people uncritically accept these type of matches as implying that they have identifiable ancestors from “exotic” places. Aside from being caused by endogamy I suppose population matches might also be a reflection of (pre)historical migration patterns. Often not verifiable by genealogy but more so correlating with rather ancient population movements across time and space.

This circumstance may hold some far-reaching implications. IBP matches are not to be dismissed in all cases. However they do require careful interpretation. In the African context countless migrations from the past may result in IBP matches. Perhaps the most impactful ones have been the Fula migrations across the Sahel corridor, from Upper Guinea into Sudan. As well as the Bantu migrations from southeast Nigeria/Cameroon into Central, East and Southern Africa.

My previous African DNA matches findings for Cape Verdeans (see this link) certainly seem to testify to the significance of the Fula migrations. The elevated frequency of presumably Hausa-Fulani matches from Nigeria perhaps being most evocative in this regard! However when performing similar surveys for other parts of the Afro-Diaspora I am quite certain that also the significance of the Bantu migrations will become apparent in matching patterns. At times confusingly so. In particular due to the appearance of unexpected matches from southern Africa (esp. Zimbabwe) as well as Kenya.

From some individual profiles I have been scanning I highly suspect that at times there may be some pile-up regions involved. Which are currently skipped by Ancestry’s Timber tool (perhaps because Timber is mostly geared towards European genetics?) I couldn’t verify this though because obviously Ancestry does not inform us about the exact location of shared DNA segments…

I have myself also observed on 23andme how Kenyans are able to receive unexpected Zimbabwean DNA matches. Seemingly due to the genetic legacy of the Bantu Expansion from many centuries or even several millennia ago. Likewise Afro-Diasporans might receive Kenyan matches due to shared Bantu origins from Central Africa (Angola/Congo). In a recently published research paper it has been revealed that Angolans were having small DNA matches with people from South Africa and also from Kenya/Uganda! See:

8) For a blog series which summarizes my take on Tracing African Roots:

9) My survey-based research has been based on the regional admixture results as well as African DNA matches being reported by both 23andme and Ancestry. See these pages for an overview:

Actually my very first survey efforts date back even earlier to 2011. Based on the pioneering African Ancestry Project by Razib Khan. I shared these findings also on 23andme’s online community at that time. They can still be seen in this online spreadsheet.

I will readily admit I am not a trained genetic genealogist.  Although I did take my first DNA test with 23andme already in 2010. And I have been keeping up with developments in the wider field of DNA testing ever since. Furthermore I do hold a university’s degree (regional economics) and therefore I am well acquainted with the basics of how scientific research should be set up as well as with scientific methodology. My research may not have been peer-reviewed. But in it self this does not say anything about the quality of my research a priori.

All of my surveys are based on empirical data, kindly shared by individual DNA testers. All of which is verifiable in online spreadsheets for which I provide links in all relevant blog posts. Therefore my data-entry & calculations can be counterchecked by anyone who feels the need to do so. On request I can also provide additional source data.

My surveys may be deemed a mere layman’s effort however my findings are not out of line with any of the published papers on the topic of Afro-Diasporan genetics I have read so far. And unlike many published studies (which are often restricted in scope) I do make an extra effort to provide as much detail and (historical) context as possible. In order to avoid oversummarization. Furthermore I also highlight individual variation and limitations of my analysis whenever I can.

16 thoughts on “Why even small African matches matter!

  1. This is an excellent post Felipe! I will definitely forward it to Ancestry. I did have one question. Have you done an African DNA Matches study for African Americans like you did for Jamaicans and if not do you plan to? I am very interested in seeing the results.

    Liked by 1 person

    • Thank you! Yes I indeed intend to do a survey on the African DNA matches being reported by Ancestry for African Americans. In fact I also am preparing similar surveys for other parts of the Afro-Diaspora. Such as West Indians (mainly Barbados, Guyana, Bahamas but also Suriname!) as well as Brazilians, Haitians, Latin Americans and Hispanic Caribbeans. So keep an eye out for that! It will take some time to process all the data though 😉

      Like

  2. oh this is not good! My mother, brother and I, have together a Dominican match. They both share with him 14 cm on 2 segments, while I share only 6 cm on 1 segment with him. I see how they wanna decrease the probability of getting false positive matches, but a good number of theses small matches are actually real matches. I believe so as I share a lot of them with both my mother and brother.

    Liked by 2 people

    • Yes, please let Ancestry know! I have seen the same for many other people whose African matches I scanned as well! And of course most people will actually not have tested their parents or other close family members. So they willl not be able to check for themselves if their smaller African matches are showing up with greater shared DNA amounts for other family members…

      Btw I recently came across this great Youtube video of an Ivorian man (Bété) talking about his African American DNA matches. Did you already see it?

      Liked by 1 person

      • Ok, I will message them, and send them the link of your blog. Hopefully they will reconsider their decision. Nope it is the first time I hear about him, thank you for sharing 🙂 i am going to check his video right now.

        Liked by 2 people

        • He has many fascinating things to say! Just like people from all around the world he has also been inspired by the original ROOTS TV series from the 70’s. He is actually a filmmaker and will also be making a documentary about his connecting with African American DNA cousins. He says he already met three of his closest matches (4th cousins, so >20 cM shared DNA!). But he has hundreds more distant matches as well.

          He has some interesting theories about how he could be related with these matches. One of these ancestral scenarios involving Kru sailors from the borderland in between Liberia and Ivory Coast. Some individuals among them might have settled in the USA during the 1800’s voluntarily he speculates. Just like Cape Verdean whalers as he mentions himself! I suppose this might account for his closest matches. However his more distant and smaller matches will probably be due to trans-atlantic slave trade.

          Around 16min he shows his African breakdown which reflects the 2018 update on Ancestry. However he also says that his original results were:

          96% “Ivory Coast/Ghana”
          3% “Hunter-Gatherers”
          <1% “Mali”

          Perfectly in line with what we have always been discussing for Kru/Bété people 😉

          Like

  3. Very interesting. I had my eyes opened with the statement of most Afro descendants trace their ancestry to the 1700 or earlier. I take great pride in the fact that I have at least 13 patriot ancestors and many of my ancestors came in the early 1600-1700s. Because I know many of my ancestors were slave owners, that also means that the slaves were also here during the same time period. I would like for each Decendant of slaves be able to have the same pride in their ancestors, knowing that they helped build this country. Not because they were slaves but because they were human beings with worth.

    Liked by 1 person

    • I agree that taking pride in our proximal and distal heritage is good. I think we need to appreciate both, which is why the current threshold should be held in tact if possible. Without our African ancestors, despite the unfortunate history of the slave trade, there would be no enslaved ancestors to revere!

      Liked by 1 person

    • Thank you for commenting! It is indeed crucial to realize that African Americans will trace back most of their African-born ancestors to the 1700’s. And therefore it is inevitable that due to dilution this 6cM-8cM range also simply happens to be the range where many of their valid African DNA matches are to be expected!

      I always find it intriguing though to see that there may be some variance according to (ultimate) state origins being either from Virginia or South Carolina. I suspect this will also influence the frequency of bigger matches or even close matches (>20cM). Which I suppose would be more so to be expected for people with South Carolina origins (on average). I intend to do more research on that in my upcoming survey on the African DNA matches being reported by Ancestry for African Americans. See also this chart taken from the Slave Voyages Database. It shows South Carolina also having a minor but significant share of African arrivals in the early 1800’s of about 30%. While for Virginia the import of African captives was mostly occurring before 1750!

      Another intriguing aspect which was heavily publicized last year is about which year would be the appropriate year to commemorate the first recorded arrival of Africans on presentday American soil. Either 1619 in Virginia or much earlier 1526 in South Carolina. See also:

      https://www.smithsonianmag.com/history/misguided-focus-1619-beginning-slavery-us-damages-our-understanding-american-history-180964873/
      https://afropunk.com/2019/09/why-1526-is-as-deserving-of-commemoration-as-1619/
      https://www.washingtonpost.com/history/2019/09/07/before-there-was-mystery-first-enslaved-africans-what-became-us/

      Like

  4. This is an excellent post, with so much takeaway! Although I struggle a bit with some of the more technical details (and charts), I’m in full agreement with the premises put forth here, and will let Ancestry know as much.
    I’ve also pulled a few quotes from you that I just love! Top of the list is this one: “… absence of evidence is not evidence of absence.” Love that!
    Thanks for your hard work and sharing.

    Renate Sanders

    Liked by 1 person

    • You’re very welcome, thanks a lot for your comment! Feel free to ask about any details or charts which need clarification. It might also be helpful for other people.

      Like

  5. Hey Fonte, Have you had a chance to look at the new paper from Micheletti?
    They’re showing that African Americans have lower Senegambian then expected from SlaveVoyages.

    Like

  6. Oh another thing. I’m wondering if the practice of polygny might not cause African centimorgans to shorten quicker. Aren’t cousins from half siblings a bit less related than cousins from full siblings? Has polygny been factored in, or are geneticists just assuming a “typical” family tree?

    Like

    • That is a good question! All I know is that this must also have had a major impact on ethnic intermingling. I have read several accounts whereby especially African men of higher social standing would often have multiple wives. Among whom would also be women from neighbouring ethnic groups. Their children would often however just assume the ethnic identities of their father. Of course also determined by localized cultural practices.

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s