October 23, 2008

Y chromosomes and mtDNA of Sweden

It's interesting to see a study of the current population of a country. If one is interested in deep pre-historical origins, one must ensure that his sample has no known foreign ancestry among the lines of interest. But, if one is interested in the current and future course of the population, then the totality of the population must be sampled.

This is a good opportunity to track shifting gene frequencies due to immigration and/or differential fertility. It would be a good idea for countries to start spending some money on a genetic census of their population. This would not need to involve all the inhabitants, and could be carried out for the fraction of cost that governments pay to collect all sorts of other statistics. Such a census would provide an important source of data to future scientists investigating the demography of Europe during this transitional era.

From the paper:
Several haplogroups had interesting frequency patterns, but also wide confidence intervals, necessitating caution in the interpretations. The mtDNA haplogroup with the strongest geographical cline, U5b, is known to have high frequencies among the northern Saami population, consistent with our results (Tambets et al. 2004). The high frequency of the Y-chromosomal R1b in the south, observed also by Karlsson et al. 2006; is consistent with its abundance in Central Europe and Denmark (Semino et al. 2000; Brion et al. 2005). Haplogroup R1a1 is more common in Norway than in Sweden (Dupuy et al. 2006), and its high frequency in Värmland/Dalarna and Halland supports the historically plausible connection to Norway (Lindqvist, 2006). Y-chromosomal haplogroup N3 (Lappalainen et al. 2006) and mtDNA haplogroup H1f (Loogväli et al. 2004; Lappalainen et al. 2008) are common in Finland, and had increased frequencies in several Swedish counties with historical ties to Finland: Eastern Svealand was the most important destination of the Finnish immigration wave in the 1970's; in Norrland the Finnish influences date back to ancient times and in Dalarna to the 17th century (Pitkänen 1994).

When compared to previous knowledge of ethnic Swedes without immigration in their familial background (Lappalainen et al. 2008), the frequencies of several haplogroups showed effects of 20th century immigration from more distant countries. The Y-chromosomal I1a had decreased frequencies in Malmö and Gothenburg most probably due to replacement by haplogroups that are common among immigrants. African immigration contributes to the frequencies of mtDNA haplogroups L3*(xN,M) and L* (xL3) (Chen et al. 2000), and Y-chromosomal haplogroup A (Underhill et al. 2001; Jobling & Tyler-Smith 2003), while Near Eastern influence can be seen in mtDNA haplogroup U7 and possibly J (Richards et al. 2000; Abu-Amero et al. 2007; Achilli et al. 2007). Asian and American immigration can be observed in the slightly elevated frequencies of mtDNA haplogroups M, A, C, D and G (Quintana-Murci et al. 2004; Hill et al. 2007) and the Y-chromosomal O, K* and P* (Underhill et al. 2001; Jobling & Tyler-Smith 2003). The frequency of the Y-chromosomal haplogroup I1b may associate to immigrants from Balkan and Eastern Europe (Rootsi et al. 2004). In Malmö and Gothenburg immigration was the main contributor to their isolated positions in the Y-chromosomal PCA plot and probably also to the higher diversities compared to the surrounding populations. These phenomena were not observed in Stockholm, where most of the immigrants come from Finland (Statistics Sweden, http://www.scb.se).

Annals of Human Genetics doi: 10.1111/j.1469-1809.2008.00487.x

Population Structure in Contemporary Sweden—A Y-Chromosomal and Mitochondrial DNA Analysis

T. Lappalainen et al.


A population sample representing the current Swedish population was analysed for maternally and paternally inherited markers with the aim of characterizing genetic variation and population structure. The sample set of 820 females and 883 males were extracted and amplified from Guthrie cards of all the children born in Sweden during one week in 2003. 14 Y-chromosomal and 34 mitochondrial DNA SNPs were genotyped. The haplogroup frequencies of the counties closest to Finland, Norway, Denmark and the Saami region in the north exhibited similarities to the neighbouring populations, resulting from the formation of the Swedish nation during the past millennium. Moreover, the recent immigration waves of the 20th century are visible in haplogroup frequencies, and have led to increased diversity and divergence of the major cities. Signs of genetic drift can be detected in several counties in northern as well as in southern Sweden. With the exception of the most drifted subpopulations, the population structure in Sweden appears mostly clinal. In conclusion, our study yielded valuable information of the structure of the Swedish population, and demonstrated the usefulness of biobanks as a source of population genetic research. Our sampling strategy, nonselective on the current population rather than stratified according to ancestry, is informative for capturing the contemporary variation in the increasingly panmictic populations of the world.



  1. I'm the first author of this paper, and I'm keeping an eye on this blog, so if anyone has any questions or comments, I'll be happy to answer.

  2. Sorry,

    I have a question. How does information of 70´s Finnish and Balkan immigrants Y-dna advance our knowledge of Swedish y-dna pool? Why such a big fuss about African, Finnish and Balkan migrants? Other than that I will thank you for your study, which I will start reading soon.

    BTW Dienekes,

    new study of Y-STR variation of Finland has come up by Palo et al, 2008. "The effect of number of loci on geographical structuring and forensic applicability of Y-STR data in Finland".

    You might want to have that appear in your blog.

  3. Tuuli,

    Any plans to go back and type your samples for more recently discovered SNPs so as to break up major haplogroups (U106 and U152 under R1b for example)?

    Additionaly at DNA Forums they have just recently (past few weeks) placed a dozen new Y SNPs under R1b and it would be interesting to see their distribution in reserch samples.


  4. BTW,

    in relation to Swedes, the new study by Palo et al, 2008, which I referred to found that the biggest Y-STR variation border in Europe is in Finland; manifested by the differences of the sampled group of Swedes in West-Coast Finland and Finns.

    "The subpopulation LMO (Swedish) differed significantly from all the
    other populations".

    "The geographical substructure among the Finnish males
    was notable when measured with the ΦST values, reaching
    values as high as ΦST=0.227 in the Yfiler data. This is
    rather extreme, given that, e.g., subpopulations Larsmo (Swedish) and
    Kymi are separated by mere 400 km, with no apparent physical dispersal barriers between them".

    Moreover, the study assessed that the most common haplotype among the sampled group of Swedes was related to I-lines.

    "From the forensic point of view,
    the substructure effects in Finland are exacerbated by the fact that clearly distinct Y-STR haplotypes dominate in different subpopulations (Fig. 4). This may be explained by gene flow from Scandinavia that occurred after the initial southeastern colonization of Finland and extended only to the southwestern parts of Finland. The eastern haplogroup N3 is the most common in Finnish males and is common in
    all parts of the country, whereas western Finland also harbor significant proportions of the Scandinavian haplogroup I1a. Indeed, the most common Y-STR
    haplotype 14-14-24-16-17-14-11.13-14-11-10-21-14-12-
    14-10-19 was confirmed to be associated with haplogroup N3 by SNP typing (data not shown). This haplotype is the most common haplotype in other subpopulations except in VA and LMO, where 14-12-23-14-16-14-14.14-13-10-10-
    22-11-11-16-10-20 and 14-13-23-16-15-14-14.14-13-10-10-
    22-11-11-16-10-20 had the highest frequencies, respectively.
    These haplotypes, associated with haplogroup I, were not encountered in most of the other subpopulations in the current sample".

  5. Richard, inferring deep ancestry of a population is not the only reason for studying population structure. Knowing the contemporary population is important e.g. for association studies of genetic causes of disease and for forensic purposes. Additionally, I think it's interesting to see how recent immigration shapes the gene pool: for example how some of the Swedish cities are diverged most probably due to immigration, but still local drift appears to have more effect in the general pattern of genetic variation in Sweden than immigration.

    test, we won't be doing any additional analysis for these samples, and at the moment I'm working with other projects that don't include Y-chromosomal or mitochondrial DNA analysis. It would have been interesting to genotype a larger set of markers, but unfortunately it wasn't possible.

    Richard, you have maybe read our paper about Y-chromosomal variation in Finland (Lappalainen et al. 2006, Gene)? The difference between eastern and western Finland is quite well documented in that article, although Palo's paper has some nice insights about forensic implications.

    If I'm allowed to advertise, take a look at PLoS ONE on Friday for a genome-wide analysis of North European populations - although I'm sure Dienekes will spot it himself.

  6. Hi Tuuli,

    Great work. In addition to genetic testing. Is there any studies on Phoenitics of Sami, Finnish, Norwegian Languages.
    can you pass any info.

    Also Is there any paper on U5b presence in Russian areas?.


  7. south central haplo,

    I'm a geneticist, so I don't really know much about phenotics references, sorry.

    U5b in Russian areas... I'm not sure and I'm too busy to check, but see Tambets et al. 2004 about Saami, they're writing quite a lot about U5b so they'll probably have references, if not actual data.

  8. Tuuli,

    Thank you for a great article.

    I am wondering about the meaning of "Regions excluding cities" in particular "SVEA2" and "GOTA2". From table 1 it appears that only the big cities Stockholm, Göteborg and Malmö have been excluded. However from the samplesizes in table 3 I can't get this to add up:
    the "components" VDA+VMA+UPP+SMA+ORE have 51 + 24 +33 +26 +32 = 166 whereas SVEA2 has a samplesize of 73. OGO+VGO+JON+HAL+SES+SKA have 37+75+28+34+39+73 = 286 whereas GOTA2 has 121. The mtdna table 2 seems to be correct. Have the big cities (such as Örebro, Jönköping etc.) been excluded in the ydna table or is there a mistake here? The frequency of P* is 8,2 in SVEA2, however according to the table no SVEA-county have a frequency of above 3,8.

    "Asian and American immigration can be observed in the slightly elevated frequencies of mtDNA haplogroups M, A, C, D and G (Quintana-Murci et al. 2004; Hill et al.2007) and the YchromosomalO, K* and P* (Underhill et al.2001; Jobling & Tyler-Smith 2003)."
    It does not seem likely to me that P*(xR1a,R1b1) in Sweden would reflect any recent immigration. In Sweden this group is almost certainly mostly Q, and Hg Q appears to have been present in Norway/Sweden for quite a long time.

    Best regards


  9. Paul,

    There's an error!! Thank you so much for pointing that out, I can't believe that me, all the other authors, reviewers and editors have missed that!

    I hope it can still be fixed before the paper's printed...

  10. Hi Tuuli!
    I just read about your fascinating dissertation topic and after googling found this site. Could you please tell me whether it is possible to get my DNA examined as to what links it has to where. My mother was born in Nuijamaa, in Karelia and my father in Alajarvi, Pohjanmaa. I'll be in Finland this summer. Who should I contact? Can your team help me? Thanks. Kindly, Heli


Stay on topic. Be polite. Use facts and arguments. Be Brief. Do not post back to back comments in the same thread, unless you absolutely have to. Don't quote excessively. Google before you ask.