05_pynb_fstatistics.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# F Statistics\n",
    "## F3 Statistics\n",
    "\n",
    "F3 statistics are a useful analytical tool to understand population relationships. F3 statistics, just as F4 and F2 statistics measure allele frequency correlations between populations and were introduced by Nick Patterson in his [Patterson 2012](http://www.genetics.org/content/early/2012/09/06/genetics.112.145037).\n",
    "\n",
    "F3 statistics are used for two purposes: i) as a test whether a target population (C) is admixed between two source populations (A and B), and ii) to measure shared drift between two test populations (A and B) from an outgroup (C).\n",
    "\n",
    "F3 statistics are in both cases defined as the product of allele frequency differences between population C to A and B, respectively:\n",
    "\n",
    "$$F3(A,B;C)=\\langle(c−a)(c−b)\\rangle$$\n",
    "\n",
    "Here, $\\langle\\cdot\\rangle$ denotes the average over all genotyped sites, and a, b and c\n",
    "denote the allele frequency for a given site in the three populations A, B and C."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Admixture F3 Statistics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It can be shown that if that statistics is negative, it provides unambiguous proof that population C is admixed between populations A and B, as in the following phylogeny (taken from Figure 1 from [Patterson 2012](http://www.genetics.org/content/early/2012/09/06/genetics.112.145037):\n",
    "\n",
    "<img src=\"img/f3_phylogeny.png\" alt=\"F3-phylogeny\" style=\"width: 300px;\"/>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Intuitively, an F3 statistics becomes negative if the allele frequency of the target population C is on average intermediate between the allele frequencies of A and B. Consider as an extreme example a genomic site where a=0, b=1 and c=0.5. Then we have (c−a)(c−b)=−0.25, which is negative. So if the entire statistics is negative, it suggests that in many positions, the allele frequency c is indeed intermediate, suggesting admixture between the two sources."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***Note:*** If an F3 statistics is *not* negative, it does *not* proof that there is no admixture!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will use this statistics to test if Finnish are admixed between East and West, using different Eastern and Western sources. In the West, we use French, Icelandic, Lithuanian and Norwegian as source, and in the East we use Nganasan and one of the ancient individuals analysed in this workshop, *Bolshoy Oleni Ostrov*, 3500 year old individuals from the Northern Russian Kola-peninsula.\n",
    "\n",
    "We use the software `qp3Pop` from [AdmixTools](https://github.com/DReichLab/AdmixTools), which similar to `smartpca` takes a parameter file:\n",
    "\n",
    "    genotypename:   input genotype file (in eigenstrat format)\n",
    "    snpname:        input snp file      (in eigenstrat format)\n",
    "    indivname:      input indiv file    (in eigenstrat format)\n",
    "    popfilename:    a file containing rows with three populations on each line A, B and C.\n",
    "    inbreed: YES"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, the last option is necessary if we are analysing pseudo-diploid ancient data (which is the case here).\n",
    "\n",
    "To prepare the `popfilename`, create a new text file using Jupyter with the following content:\n",
    "\n",
    "    Nganasan French Finnish\n",
    "\tNganasan Icelandic Finnish\n",
    "\tNganasan Lithuanian Finnish\n",
    "\tNganasan Norwegian Finnish\n",
    "\tBolshoyOleniOstrov French Finnish\n",
    "\tBolshoyOleniOstrov Icelandic Finnish\n",
    "\tBolshoyOleniOstrov Lithuanian Finnish\n",
    "\tBolshoyOleniOstrov Norwegian Finnish"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***Exercise:*** Prepare the parameter file with the input data as in the PCA session (see Principal Components Analysis (PCA)) and then run `qp3Pop -p PARAMETER_FILE`, where `PARAMETERFILE` should be replaced by your parameter file name. As genotype data, use the files called `/data/popgen_course/HumanOrigins_FennoScandian_small.*`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The results are in the output that you can view in the Notebook. The crucial bit should look like this:\n",
    "\n",
    "\tSource 1\tSource 2\tTarget\tf_3\tstd. err\tZ\tSNPs\n",
    "\tresult:\tNganasan\tFrench\tFinnish\t-0.004539\t0.000510\t-8.894\t442567\n",
    "\tresult:\tNganasan\tIcelandic\tFinnish\t-0.005297\t0.000563\t-9.404\t427954\n",
    "\tresult:\tNganasan\tLithuanian\tFinnish\t-0.005062\t0.000590\t-8.574\t426231\n",
    "\tresult:\tNganasan\tNorwegian\tFinnish\t-0.004744\t0.000569\t-8.332\t428161\n",
    "\tresult:\tBolshoyOleniOstrov\tFrench\tFinnish\t-0.002814\t0.000444\t-6.341\t402958\n",
    "\tresult:\tBolshoyOleniOstrov\tIcelandic\tFinnish\t-0.002590\t0.000486\t-5.323\t386418\n",
    "\tresult:\tBolshoyOleniOstrov\tLithuanian\tFinnish\t-0.001523\t0.000536\t-2.840\t384134\n",
    "\tresult:\tBolshoyOleniOstrov\tNorwegian\tFinnish\t-0.001553\t0.000502\t-3.092\t386203\n",
    "\n",
    "This output shows as first three columns the three populations A, B (sources) and C (target). Then the f3 statistics, which is negative in all cases tested here, a standard error, a Z score and the number of SNPs involved in the statistics.\n",
    "\n",
    "The Z score is key: It gives the deviation of the f3 statistic from zero in units of the standard error. As general rule, a Z score of -3 or more suggests a significant rejection of the Null hypothesis that the statistic is not negative. In this case, all of the statistics are significantly negative, proving that Finnish have ancestral admixture of East and West Eurasian ancestry. Note that the statistics does not suggest when this admixture happened!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## F4 Statistics\n",
    "\n",
    "A different way to test for admixture is by “F4 statistics” (or “D statistics” which is very similar), also introduced in [Patterson 2012](http://www.genetics.org/content/early/2012/09/06/genetics.112.145037).\n",
    "\n",
    "F4 statistics are also defined in terms of correlations of allele frequency differences, similarly to F3 statistics (see above), but involving four different populations, not just three. Specifically we define\n",
    "\n",
    "$$F4(A,B;C,D)=\\langle(a−b)(c−d)\\rangle.$$\n",
    "\n",
    "To understand the statistics, consider the following tree:\n",
    "\n",
    "<img src=\"img/f4_phylogeny.png\" alt=\"F4-phylogeny\" style=\"width: 300px;\"/>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this tree, without any additional admixture, the allele frequency difference between A and B should be completely independent from the allele frequency difference between C and D. In that case, F4(A, B; C, D) should be zero, or at least not statistically different from zero. However, if there was gene flow from C or D into A or B, the statistic should be different from zero. Specifically, if the statistic is significantly negative, it implies gene flow between either C and B, or D and A. If it is significantly positive, it implies gene flow between A and C, or B and D.\n",
    "\n",
    "The way this statistic is often used, is to put a divergent outgroup as population A, for which we know for sure that there was no admixture into either C or D. With this setup, we can then test for gene flow between B and D (if the statistic is positive), or B and C (if it is negative).\n",
    "\n",
    "Here, we can use this statistic to test for East Asian admixture in Finns, similarly to the test using Admixture F3 statistics above. We will use the `qpDstat` program from [AdmixTools](https://github.com/DReichLab/AdmixTools) for that. We need to again prepare a population list file, this time with four populations (A, B, C, D). I suggest you open a new file and fill it with:\n",
    "\n",
    "\tMbuti Nganasan French Finnish\n",
    "\tMbuti Nganasan Icelandic Finnish\n",
    "\tMbuti Nganasan Lithuanian Finnish\n",
    "\tMbuti Nganasan Norwegian Finnish\n",
    "\tMbuti BolshoyOleniOstrov French Finnish\n",
    "\tMbuti BolshoyOleniOstrov Icelandic Finnish\n",
    "\tMbuti BolshoyOleniOstrov Lithuanian Finnish\n",
    "\tMbuti BolshoyOleniOstrov Norwegian Finnish"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can then use this file again in a parameter file, similar to the one prepared for `qp3Pop` above:\n",
    "\n",
    "\tgenotypename:   input genotype file (in eigenstrat format)\n",
    "\tsnpname:        input snp file      (in eigenstrat format)\n",
    "\tindivname:      input indiv file    (in eigenstrat format)\n",
    "\tpopfilename:    a file containing rows with three populations on each line A, B and C.\n",
    "\tf4mode: YES\n",
    "\n",
    "Note that you cannot give the “inbreed” option here."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***Exercise:*** Prepare the parameter file as suggested above and then run `qpDstat -p PARAMETER_FILE`, where `PARAMETERFILE` should be replaced by your parameter file name. This will take 5-6 minutes."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The results should be (skipping some header lines):\n",
    "\n",
    "    result:\tMbuti\tNganasan     French    Finnish      0.002363     19.016   29254  27852 593124\n",
    "\tresult:\tMbuti\tNganasan  Icelandic    Finnish      0.001721     11.926   28915  27894 593124\n",
    "\tresult:\tMbuti\tNganasan Lithuanian    Finnish      0.001368      9.664   28745  27933 593124\n",
    "\tresult:\tMbuti\tNganasan  Norwegian    Finnish      0.001685     11.663   28933  27934 593124\n",
    "\tresult:\tMbuti\tBolshoyOleniOstrov     French    Finnish      0.001962     16.737   27249  26175 547486\n",
    "\tresult:\tMbuti\tBolshoyOleniOstrov  Icelandic    Finnish      0.001084      7.776   26876  26282 547486\n",
    "\tresult:\tMbuti\tBolshoyOleniOstrov Lithuanian    Finnish      0.000554      3.942   26683  26380 547486\n",
    "\tresult:\tMbuti\tBolshoyOleniOstrov  Norwegian    Finnish      0.000952      6.707   26873  26351 547486"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, the key columns are columns 2, 3, 4 and 5, denoting A, B, C and D, and column 6 and 7, which denote the F4 statistic and the Z score, measuring significance in difference from zero.\n",
    "\n",
    "As you can see, in all cases, the Z score is positive and larger than 3, indicating a significant deviation from zero, and implying gene flow between Nganasan and Finnish, and BolshoyOleniOstrov and Finnish, when compared to French, Icelandic, Lithuanian or Norwegian."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Outgroup-F3-Statistics\n",
    "Outgroup F3 statistics are a special case how to use F3 statistics. The definition is the same as for Admixture F3 statistics, but instead of a target C and two source populations A and B, one now gives an outgroup C and two test populations A and B.\n",
    "\n",
    "To get an intuition for this statistics, consider the following tree:\n",
    "\n",
    "<img src=\"img/outgroupf3_phylogeny.png\" alt=\"Outgroup-F3-phylogeny\" style=\"width: 300px;\"/>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this scenario, the statistic F3(A, B; C) measures the branch length from C to the common ancestor of A and B, coloured red. So this statistic is simply a measure of how closely two population A and B are related with each other, as measured from a distant outgroup. It is thus a similarity measure: The higher the statistic, the more genetically similar A and B are to one another.\n",
    "\n",
    "We can use this statistic to measure for example the the genetic affinity to East Asia, by performing the statistic F3(Han, X; Mbuti), where Mbuti is a distant African population and acts as outgroup here, Han denote Han Chinese, and X denotes various European populations that we want to test.\n",
    "\n",
    "You need to start, again, by preparing a list of population triples to be measured. I suggest the following list:\n",
    "\n",
    "\tHan Chuvash Mbuti\n",
    "\tHan Albanian Mbuti\n",
    "\tHan Armenian Mbuti\n",
    "\tHan Bulgarian Mbuti\n",
    "\tHan Czech Mbuti\n",
    "\tHan Druze Mbuti\n",
    "\tHan English Mbuti\n",
    "\tHan Estonian Mbuti\n",
    "\tHan Finnish Mbuti\n",
    "\tHan French Mbuti\n",
    "\tHan Georgian Mbuti\n",
    "\tHan Greek Mbuti\n",
    "\tHan Hungarian Mbuti\n",
    "\tHan Icelandic Mbuti\n",
    "\tHan Italian_North Mbuti\n",
    "\tHan Italian_South Mbuti\n",
    "\tHan Lithuanian Mbuti\n",
    "\tHan Maltese Mbuti\n",
    "\tHan Mordovian Mbuti\n",
    "\tHan Norwegian Mbuti\n",
    "\tHan Orcadian Mbuti\n",
    "\tHan Russian Mbuti\n",
    "\tHan Sardinian Mbuti\n",
    "\tHan Scottish Mbuti\n",
    "\tHan Sicilian Mbuti\n",
    "\tHan Spanish_North Mbuti\n",
    "\tHan Spanish Mbuti\n",
    "\tHan Ukrainian Mbuti\n",
    "\tHan Levanluhta Mbuti\n",
    "\tHan BolshoyOleniOstrov Mbuti\n",
    "\tHan ChalmnyVarre Mbuti\n",
    "\tHan Saami.DG Mbuti"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "which cycles through many populations from Europe, including the ancient individuals from Chalmny Varre, Bolshoy Oleni Ostrov and Levänluhta."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***Exercise:*** Copy this list into a file, and prepare a parameter file for running `qp3Pop`, similar to the parameter file for admixture F3 statistics above, and run `qp3Pop` with that parameter file as above. Note that here you don't need the line beginning with `inbreed`. This will take up to 10 minutes."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You should find this (skipping header lines from the output):\n",
    "\n",
    "\t           Source 1             Source 2   Target        f_3       std. err           Z    SNPs\n",
    "\tresult:        Han              Chuvash    Mbuti   0.233652       0.002072     112.782  502678\n",
    "\tresult:        Han             Albanian    Mbuti   0.215629       0.002029     106.291  501734\n",
    "\tresult:        Han             Armenian    Mbuti   0.213724       0.001963     108.882  504370\n",
    "\tresult:        Han            Bulgarian    Mbuti   0.216193       0.001979     109.266  504310\n",
    "\tresult:        Han                Czech    Mbuti   0.218060       0.002002     108.939  504089\n",
    "\tresult:        Han                Druze    Mbuti   0.209551       0.001919     109.205  510853\n",
    "\tresult:        Han              English    Mbuti   0.216959       0.001973     109.954  504161\n",
    "\tresult:        Han             Estonian    Mbuti   0.220730       0.002019     109.332  503503\n",
    "\tresult:        Han              Finnish    Mbuti   0.223447       0.002044     109.345  502217\n",
    "\tresult:        Han               French    Mbuti   0.216623       0.001969     110.012  509613\n",
    "\tresult:        Han             Georgian    Mbuti   0.214295       0.001935     110.721  503598\n",
    "\tresult:        Han                Greek    Mbuti   0.215203       0.001984     108.465  507475\n",
    "\tresult:        Han            Hungarian    Mbuti   0.217894       0.001999     109.004  507409\n",
    "\tresult:        Han            Icelandic    Mbuti   0.218683       0.002015     108.553  504655\n",
    "\tresult:        Han        Italian_North    Mbuti   0.215332       0.001978     108.854  507589\n",
    "\tresult:        Han        Italian_South    Mbuti   0.211787       0.002271      93.265  492400\n",
    "\tresult:        Han           Lithuanian    Mbuti   0.219615       0.002032     108.098  503681\n",
    "\tresult:        Han              Maltese    Mbuti   0.210359       0.001956     107.542  503985\n",
    "\tresult:        Han            Mordovian    Mbuti   0.223469       0.002008     111.296  503441\n",
    "\tresult:        Han            Norwegian    Mbuti   0.218873       0.002023     108.197  504621\n",
    "\tresult:        Han             Orcadian    Mbuti   0.217773       0.002014     108.115  504993\n",
    "\tresult:        Han              Russian    Mbuti   0.223993       0.001995     112.274  506525\n",
    "\tresult:        Han            Sardinian    Mbuti   0.213230       0.001980     107.711  508413\n",
    "\tresult:        Han             Scottish    Mbuti   0.218489       0.002039     107.145  499784\n",
    "\tresult:        Han             Sicilian    Mbuti   0.212272       0.001975     107.486  505477\n",
    "\tresult:        Han        Spanish_North    Mbuti   0.215885       0.002029     106.383  500853\n",
    "\tresult:        Han              Spanish    Mbuti   0.213869       0.001975     108.297  513648\n",
    "\tresult:        Han            Ukrainian    Mbuti   0.218716       0.002007     108.950  503981\n",
    "\tresult:        Han           Levanluhta    Mbuti   0.236252       0.002383      99.123  263049\n",
    "\tresult:        Han   BolshoyOleniOstrov    Mbuti   0.247814       0.002177     113.849  457102\n",
    "\tresult:        Han         ChalmnyVarre    Mbuti   0.233499       0.002304     101.345  366220\n",
    "\tresult:        Han             Saami.DG    Mbuti   0.236198       0.002274     103.852  489038"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now it’s time to plot these results using python. Copy the results (all lines from the output beginning with “results:”) into a text file named \"f3_outgroup_stats_Han.txt\", and load it into a pandas dataframe using:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "d = pd.read_csv(\"f3_outgroup_stats_Han.txt\",\n",
    "                delim_whitespace=True,\n",
    "                names=[\"dummy\", \"A\", \"B\", \"C\", \"F3\", \"StdErr\", \"Z\", \"SNPS\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can check that it worked:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>dummy</th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>F3</th>\n",
       "      <th>StdErr</th>\n",
       "      <th>Z</th>\n",
       "      <th>SNPS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Chuvash</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.233652</td>\n",
       "      <td>0.002072</td>\n",
       "      <td>112.782</td>\n",
       "      <td>502678</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Albanian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.215629</td>\n",
       "      <td>0.002029</td>\n",
       "      <td>106.291</td>\n",
       "      <td>501734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Armenian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.213724</td>\n",
       "      <td>0.001963</td>\n",
       "      <td>108.882</td>\n",
       "      <td>504370</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Bulgarian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.216193</td>\n",
       "      <td>0.001979</td>\n",
       "      <td>109.266</td>\n",
       "      <td>504310</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Czech</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.218060</td>\n",
       "      <td>0.002002</td>\n",
       "      <td>108.939</td>\n",
       "      <td>504089</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Druze</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.209551</td>\n",
       "      <td>0.001919</td>\n",
       "      <td>109.205</td>\n",
       "      <td>510853</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>English</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.216959</td>\n",
       "      <td>0.001973</td>\n",
       "      <td>109.954</td>\n",
       "      <td>504161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Estonian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.220730</td>\n",
       "      <td>0.002019</td>\n",
       "      <td>109.332</td>\n",
       "      <td>503503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Finnish</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.223447</td>\n",
       "      <td>0.002044</td>\n",
       "      <td>109.345</td>\n",
       "      <td>502217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>French</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.216623</td>\n",
       "      <td>0.001969</td>\n",
       "      <td>110.012</td>\n",
       "      <td>509613</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Georgian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.214295</td>\n",
       "      <td>0.001935</td>\n",
       "      <td>110.721</td>\n",
       "      <td>503598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Greek</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.215203</td>\n",
       "      <td>0.001984</td>\n",
       "      <td>108.465</td>\n",
       "      <td>507475</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Hungarian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.217894</td>\n",
       "      <td>0.001999</td>\n",
       "      <td>109.004</td>\n",
       "      <td>507409</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Icelandic</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.218683</td>\n",
       "      <td>0.002015</td>\n",
       "      <td>108.553</td>\n",
       "      <td>504655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Italian_North</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.215332</td>\n",
       "      <td>0.001978</td>\n",
       "      <td>108.854</td>\n",
       "      <td>507589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Italian_South</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.211787</td>\n",
       "      <td>0.002271</td>\n",
       "      <td>93.265</td>\n",
       "      <td>492400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Lithuanian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.219615</td>\n",
       "      <td>0.002032</td>\n",
       "      <td>108.098</td>\n",
       "      <td>503681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Maltese</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.210359</td>\n",
       "      <td>0.001956</td>\n",
       "      <td>107.542</td>\n",
       "      <td>503985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Mordovian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.223469</td>\n",
       "      <td>0.002008</td>\n",
       "      <td>111.296</td>\n",
       "      <td>503441</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Norwegian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.218873</td>\n",
       "      <td>0.002023</td>\n",
       "      <td>108.197</td>\n",
       "      <td>504621</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Orcadian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.217773</td>\n",
       "      <td>0.002014</td>\n",
       "      <td>108.115</td>\n",
       "      <td>504993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Russian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.223993</td>\n",
       "      <td>0.001995</td>\n",
       "      <td>112.274</td>\n",
       "      <td>506525</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Sardinian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.213230</td>\n",
       "      <td>0.001980</td>\n",
       "      <td>107.711</td>\n",
       "      <td>508413</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Scottish</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.218489</td>\n",
       "      <td>0.002039</td>\n",
       "      <td>107.145</td>\n",
       "      <td>499784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Sicilian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.212272</td>\n",
       "      <td>0.001975</td>\n",
       "      <td>107.486</td>\n",
       "      <td>505477</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Spanish_North</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.215885</td>\n",
       "      <td>0.002029</td>\n",
       "      <td>106.383</td>\n",
       "      <td>500853</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Spanish</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.213869</td>\n",
       "      <td>0.001975</td>\n",
       "      <td>108.297</td>\n",
       "      <td>513648</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Ukrainian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.218716</td>\n",
       "      <td>0.002007</td>\n",
       "      <td>108.950</td>\n",
       "      <td>503981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Levanluhta</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.236252</td>\n",
       "      <td>0.002383</td>\n",
       "      <td>99.123</td>\n",
       "      <td>263049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>BolshoyOleniOstrov</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.247814</td>\n",
       "      <td>0.002177</td>\n",
       "      <td>113.849</td>\n",
       "      <td>457102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>ChalmnyVarre</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.233499</td>\n",
       "      <td>0.002304</td>\n",
       "      <td>101.345</td>\n",
       "      <td>366220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Saami.DG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.236198</td>\n",
       "      <td>0.002274</td>\n",
       "      <td>103.852</td>\n",
       "      <td>489038</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      dummy    A                   B      C        F3    StdErr        Z  \\\n",
       "0   result:  Han             Chuvash  Mbuti  0.233652  0.002072  112.782   \n",
       "1   result:  Han            Albanian  Mbuti  0.215629  0.002029  106.291   \n",
       "2   result:  Han            Armenian  Mbuti  0.213724  0.001963  108.882   \n",
       "3   result:  Han           Bulgarian  Mbuti  0.216193  0.001979  109.266   \n",
       "4   result:  Han               Czech  Mbuti  0.218060  0.002002  108.939   \n",
       "5   result:  Han               Druze  Mbuti  0.209551  0.001919  109.205   \n",
       "6   result:  Han             English  Mbuti  0.216959  0.001973  109.954   \n",
       "7   result:  Han            Estonian  Mbuti  0.220730  0.002019  109.332   \n",
       "8   result:  Han             Finnish  Mbuti  0.223447  0.002044  109.345   \n",
       "9   result:  Han              French  Mbuti  0.216623  0.001969  110.012   \n",
       "10  result:  Han            Georgian  Mbuti  0.214295  0.001935  110.721   \n",
       "11  result:  Han               Greek  Mbuti  0.215203  0.001984  108.465   \n",
       "12  result:  Han           Hungarian  Mbuti  0.217894  0.001999  109.004   \n",
       "13  result:  Han           Icelandic  Mbuti  0.218683  0.002015  108.553   \n",
       "14  result:  Han       Italian_North  Mbuti  0.215332  0.001978  108.854   \n",
       "15  result:  Han       Italian_South  Mbuti  0.211787  0.002271   93.265   \n",
       "16  result:  Han          Lithuanian  Mbuti  0.219615  0.002032  108.098   \n",
       "17  result:  Han             Maltese  Mbuti  0.210359  0.001956  107.542   \n",
       "18  result:  Han           Mordovian  Mbuti  0.223469  0.002008  111.296   \n",
       "19  result:  Han           Norwegian  Mbuti  0.218873  0.002023  108.197   \n",
       "20  result:  Han            Orcadian  Mbuti  0.217773  0.002014  108.115   \n",
       "21  result:  Han             Russian  Mbuti  0.223993  0.001995  112.274   \n",
       "22  result:  Han           Sardinian  Mbuti  0.213230  0.001980  107.711   \n",
       "23  result:  Han            Scottish  Mbuti  0.218489  0.002039  107.145   \n",
       "24  result:  Han            Sicilian  Mbuti  0.212272  0.001975  107.486   \n",
       "25  result:  Han       Spanish_North  Mbuti  0.215885  0.002029  106.383   \n",
       "26  result:  Han             Spanish  Mbuti  0.213869  0.001975  108.297   \n",
       "27  result:  Han           Ukrainian  Mbuti  0.218716  0.002007  108.950   \n",
       "28  result:  Han          Levanluhta  Mbuti  0.236252  0.002383   99.123   \n",
       "29  result:  Han  BolshoyOleniOstrov  Mbuti  0.247814  0.002177  113.849   \n",
       "30  result:  Han        ChalmnyVarre  Mbuti  0.233499  0.002304  101.345   \n",
       "31  result:  Han            Saami.DG  Mbuti  0.236198  0.002274  103.852   \n",
       "\n",
       "      SNPS  \n",
       "0   502678  \n",
       "1   501734  \n",
       "2   504370  \n",
       "3   504310  \n",
       "4   504089  \n",
       "5   510853  \n",
       "6   504161  \n",
       "7   503503  \n",
       "8   502217  \n",
       "9   509613  \n",
       "10  503598  \n",
       "11  507475  \n",
       "12  507409  \n",
       "13  504655  \n",
       "14  507589  \n",
       "15  492400  \n",
       "16  503681  \n",
       "17  503985  \n",
       "18  503441  \n",
       "19  504621  \n",
       "20  504993  \n",
       "21  506525  \n",
       "22  508413  \n",
       "23  499784  \n",
       "24  505477  \n",
       "25  500853  \n",
       "26  513648  \n",
       "27  503981  \n",
       "28  263049  \n",
       "29  457102  \n",
       "30  366220  \n",
       "31  489038  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "d_sorted = d.sort_values(by=\"F3\")\n",
    "y = range(len(d_sorted))\n",
    "plt.figure(figsize=(6, 8))\n",
    "plt.errorbar(d_sorted[\"F3\"], y, xerr=d[\"StdErr\"], fmt='o')\n",
    "plt.yticks(y, d_sorted[\"B\"]);\n",
    "plt.xlabel(\"F3(Han, Test; Mbuti)\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As expected, the ancient samples and modern Saami are the ones with the highest allele sharing with present-day East Asians (as represented by Han) compared to many other Europeans."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Outgroup F3 Statistics Scatter plot\n",
    "\n",
    "The above plot shows an intriguing cline of differential relatedness to Han in many Europeans. For example, would you have guessed that Icelandics are closer to Han than Armenians are to Han? This is very surprising, and it shows that European ancestry has a complex relationship to East Asians. To understand this better, you can read [Patterson 2012](http://www.genetics.org/content/early/2012/09/06/genetics.112.145037), who makes some intriguing observations. Patterson and colleagues use Admixture F3 statistics and apply it to many populations world-wide. They summarise some population triples with the most negative F3 statistics in the following table:\n",
    "\n",
    "<img src=\"img/Patterson_2012_table.png\" alt=\"Patterson 2012 table\" style=\"width: 450px;\"/>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are many interesting results here, but one of the most striking one is the finding of F3(Sardinian, Karitiana; French), which is highly significantly negative. This statistics implies that French are admixed between Sardinians and Karitiana, a Native American population from Brazil. How is that possible? We can of course rule out any recent Native American backflow into Europe.\n",
    "\n",
    "Patterson and colleagues explained this finding with hypothesising an ancient admixture event, from a Siberian population that contributed to both Europeans and to Native Americans. They termed that population the “Ancient North Eurasians (ANE)”. The following admixture graph was suggested:\n",
    "\n",
    "<img src=\"img/Patterson_2012_ANEfig.png\" alt=\"Patterson 2012 ANE graph\" style=\"width: 600px;\"/>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see, the idea is that modern Central Europeans, such as French, are admixed between Southern Europeans (Sardinians) and ANE. The Ancient North Eurasians are a classic example for a “Ghost” population, a population which does not exist anymore in unmixed form, and from which we have no direct individual representative.\n",
    "\n",
    "Amazingly, two years after the publication of [Patterson 2012](http://www.genetics.org/content/early/2012/09/06/genetics.112.145037), the ANE ghost population was actually found: [Raghavan et al.](https://www.nature.com/articles/nature12736) and colleagues, in 2014, published a paper called “Upper Palaeolithic Siberian genome reveals dual ancestry of Native Americans”. A 24,000 year old boy (called MA1) from the site of “Mal’ta” in Siberia was shown to have close genetic affinity with both Europeans and in particular Native Americans, just as proposed in [Patterson 2012](http://www.genetics.org/content/early/2012/09/06/genetics.112.145037).\n",
    "\n",
    "The affinities are summarised nicely in this figure from [Raghavan et al.](https://www.nature.com/articles/nature12736):\n",
    "\n",
    "<img src=\"img/MA1_affinities.png\" alt=\"MA1 Affinities\" style=\"width: 600px;\"/>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "OK, so we now know that ancestry related to Native Americans contributed to European countries. Could that possibly explain the affinity of our ancient samples and Saami to Han Chinese in some way? To test this, we will run the same Outgroup F3 statistics as above, but this time not with Han but with MA1 as test population. Specifically, we run the following population triples in `qp3Pop`:\n",
    "\n",
    "\tMA1_HG.SG Chuvash Mbuti\n",
    "\tMA1_HG.SG Albanian Mbuti\n",
    "\tMA1_HG.SG Armenian Mbuti\n",
    "\tMA1_HG.SG Bulgarian Mbuti\n",
    "\tMA1_HG.SG Czech Mbuti\n",
    "\tMA1_HG.SG Druze Mbuti\n",
    "\tMA1_HG.SG English Mbuti\n",
    "\tMA1_HG.SG Estonian Mbuti\n",
    "\tMA1_HG.SG Finnish Mbuti\n",
    "\tMA1_HG.SG French Mbuti\n",
    "\tMA1_HG.SG Georgian Mbuti\n",
    "\tMA1_HG.SG Greek Mbuti\n",
    "\tMA1_HG.SG Hungarian Mbuti\n",
    "\tMA1_HG.SG Icelandic Mbuti\n",
    "\tMA1_HG.SG Italian_North Mbuti\n",
    "\tMA1_HG.SG Italian_South Mbuti\n",
    "\tMA1_HG.SG Lithuanian Mbuti\n",
    "\tMA1_HG.SG Maltese Mbuti\n",
    "\tMA1_HG.SG Mordovian Mbuti\n",
    "\tMA1_HG.SG Norwegian Mbuti\n",
    "\tMA1_HG.SG Orcadian Mbuti\n",
    "\tMA1_HG.SG Russian Mbuti\n",
    "\tMA1_HG.SG Sardinian Mbuti\n",
    "\tMA1_HG.SG Scottish Mbuti\n",
    "\tMA1_HG.SG Sicilian Mbuti\n",
    "\tMA1_HG.SG Spanish_North Mbuti\n",
    "\tMA1_HG.SG Spanish Mbuti\n",
    "\tMA1_HG.SG Ukrainian Mbuti\n",
    "\tMA1_HG.SG Levanluhta Mbuti\n",
    "\tMA1_HG.SG BolshoyOleniOstrov Mbuti\n",
    "\tMA1_HG.SG ChalmnyVarre Mbuti\n",
    "\tMA1_HG.SG Saami.DG Mbuti\n",
    "\n",
    "here, `MA1_HG.SG` is the somewhat cryptic population name for the MA1-individual."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***Exercise:*** Follow the same protocol as above: Copy the list into a file, prepare a parameter file for `qp3Pop` with that population triple list, and run `qp3Pop`. Copy the results (all lines beginning with “results:”) into a file, named \"f3_outgroup_stats_MA1.txt\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To test in what way the relationship to Han Chinese is correlated with the relationship with MA1, we will now plot the two statistics against each other in a scatter plot. We first have to merge the two outgroup-F3 datasets together. Here is the code including loading (assuming that the two F3 dataframes are called `outgroupf3dat_Han` and `outgroupf3dat_MA1`):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "outgroupf3dat_Han = pd.read_csv(\"f3_outgroup_stats_Han.txt\",\n",
    "                        delim_whitespace=True,\n",
    "                        names=[\"dummy\", \"A\", \"B\", \"C\", \"F3\", \"stderr\", \"Z\", \"nSNPs\"])\n",
    "outgroupf3dat_MA1 = pd.read_csv(\"f3_outgroup_stats_MA1.txt\",\n",
    "                        delim_whitespace=True,\n",
    "                        names=[\"dummy\", \"A\", \"B\", \"C\", \"F3\", \"stderr\", \"Z\", \"nSNPs\"])\n",
    "\n",
    "outgroupf3dat_merged = outgroupf3dat_Han.merge(outgroupf3dat_MA1, on=\"B\", suffixes=(\"_Han\", \"_MA1\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Again, we check that everything worked:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>dummy_Han</th>\n",
       "      <th>A_Han</th>\n",
       "      <th>B</th>\n",
       "      <th>C_Han</th>\n",
       "      <th>F3_Han</th>\n",
       "      <th>stderr_Han</th>\n",
       "      <th>Z_Han</th>\n",
       "      <th>nSNPs_Han</th>\n",
       "      <th>dummy_MA1</th>\n",
       "      <th>A_MA1</th>\n",
       "      <th>C_MA1</th>\n",
       "      <th>F3_MA1</th>\n",
       "      <th>stderr_MA1</th>\n",
       "      <th>Z_MA1</th>\n",
       "      <th>nSNPs_MA1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Chuvash</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.233652</td>\n",
       "      <td>0.002072</td>\n",
       "      <td>112.782</td>\n",
       "      <td>502678</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.243818</td>\n",
       "      <td>0.002349</td>\n",
       "      <td>103.781</td>\n",
       "      <td>350484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Albanian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.215629</td>\n",
       "      <td>0.002029</td>\n",
       "      <td>106.291</td>\n",
       "      <td>501734</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.236494</td>\n",
       "      <td>0.002296</td>\n",
       "      <td>103.008</td>\n",
       "      <td>344332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Armenian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.213724</td>\n",
       "      <td>0.001963</td>\n",
       "      <td>108.882</td>\n",
       "      <td>504370</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.231399</td>\n",
       "      <td>0.002264</td>\n",
       "      <td>102.229</td>\n",
       "      <td>349612</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Bulgarian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.216193</td>\n",
       "      <td>0.001979</td>\n",
       "      <td>109.266</td>\n",
       "      <td>504310</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.237498</td>\n",
       "      <td>0.002281</td>\n",
       "      <td>104.103</td>\n",
       "      <td>349800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Czech</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.218060</td>\n",
       "      <td>0.002002</td>\n",
       "      <td>108.939</td>\n",
       "      <td>504089</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.243224</td>\n",
       "      <td>0.002328</td>\n",
       "      <td>104.457</td>\n",
       "      <td>349553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Druze</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.209551</td>\n",
       "      <td>0.001919</td>\n",
       "      <td>109.205</td>\n",
       "      <td>510853</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.226740</td>\n",
       "      <td>0.002197</td>\n",
       "      <td>103.193</td>\n",
       "      <td>359004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>English</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.216959</td>\n",
       "      <td>0.001973</td>\n",
       "      <td>109.954</td>\n",
       "      <td>504161</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.243135</td>\n",
       "      <td>0.002317</td>\n",
       "      <td>104.941</td>\n",
       "      <td>349321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Estonian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.220730</td>\n",
       "      <td>0.002019</td>\n",
       "      <td>109.332</td>\n",
       "      <td>503503</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.247065</td>\n",
       "      <td>0.002362</td>\n",
       "      <td>104.619</td>\n",
       "      <td>348861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Finnish</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.223447</td>\n",
       "      <td>0.002044</td>\n",
       "      <td>109.345</td>\n",
       "      <td>502217</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.245684</td>\n",
       "      <td>0.002379</td>\n",
       "      <td>103.266</td>\n",
       "      <td>347208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>French</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.216623</td>\n",
       "      <td>0.001969</td>\n",
       "      <td>110.012</td>\n",
       "      <td>509613</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.240235</td>\n",
       "      <td>0.002269</td>\n",
       "      <td>105.886</td>\n",
       "      <td>357842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Georgian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.214295</td>\n",
       "      <td>0.001935</td>\n",
       "      <td>110.721</td>\n",
       "      <td>503598</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.232645</td>\n",
       "      <td>0.002253</td>\n",
       "      <td>103.243</td>\n",
       "      <td>349082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Greek</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.215203</td>\n",
       "      <td>0.001984</td>\n",
       "      <td>108.465</td>\n",
       "      <td>507475</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.236566</td>\n",
       "      <td>0.002280</td>\n",
       "      <td>103.757</td>\n",
       "      <td>355261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Hungarian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.217894</td>\n",
       "      <td>0.001999</td>\n",
       "      <td>109.004</td>\n",
       "      <td>507409</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.241720</td>\n",
       "      <td>0.002313</td>\n",
       "      <td>104.483</td>\n",
       "      <td>355340</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Icelandic</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.218683</td>\n",
       "      <td>0.002015</td>\n",
       "      <td>108.553</td>\n",
       "      <td>504655</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.244488</td>\n",
       "      <td>0.002386</td>\n",
       "      <td>102.481</td>\n",
       "      <td>350287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Italian_North</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.215332</td>\n",
       "      <td>0.001978</td>\n",
       "      <td>108.854</td>\n",
       "      <td>507589</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.236407</td>\n",
       "      <td>0.002273</td>\n",
       "      <td>104.002</td>\n",
       "      <td>354999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Italian_South</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.211787</td>\n",
       "      <td>0.002271</td>\n",
       "      <td>93.265</td>\n",
       "      <td>492400</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.230839</td>\n",
       "      <td>0.002767</td>\n",
       "      <td>83.427</td>\n",
       "      <td>321217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Lithuanian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.219615</td>\n",
       "      <td>0.002032</td>\n",
       "      <td>108.098</td>\n",
       "      <td>503681</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.246864</td>\n",
       "      <td>0.002403</td>\n",
       "      <td>102.718</td>\n",
       "      <td>348656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Maltese</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.210359</td>\n",
       "      <td>0.001956</td>\n",
       "      <td>107.542</td>\n",
       "      <td>503985</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.230200</td>\n",
       "      <td>0.002259</td>\n",
       "      <td>101.903</td>\n",
       "      <td>347725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Mordovian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.223469</td>\n",
       "      <td>0.002008</td>\n",
       "      <td>111.296</td>\n",
       "      <td>503441</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.245284</td>\n",
       "      <td>0.002346</td>\n",
       "      <td>104.571</td>\n",
       "      <td>350058</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Norwegian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.218873</td>\n",
       "      <td>0.002023</td>\n",
       "      <td>108.197</td>\n",
       "      <td>504621</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.243930</td>\n",
       "      <td>0.002301</td>\n",
       "      <td>106.031</td>\n",
       "      <td>350182</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Orcadian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.217773</td>\n",
       "      <td>0.002014</td>\n",
       "      <td>108.115</td>\n",
       "      <td>504993</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.243614</td>\n",
       "      <td>0.002320</td>\n",
       "      <td>105.008</td>\n",
       "      <td>351053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Russian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.223993</td>\n",
       "      <td>0.001995</td>\n",
       "      <td>112.274</td>\n",
       "      <td>506525</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.245212</td>\n",
       "      <td>0.002298</td>\n",
       "      <td>106.698</td>\n",
       "      <td>355953</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Sardinian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.213230</td>\n",
       "      <td>0.001980</td>\n",
       "      <td>107.711</td>\n",
       "      <td>508413</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.231967</td>\n",
       "      <td>0.002264</td>\n",
       "      <td>102.449</td>\n",
       "      <td>355548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Scottish</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.218489</td>\n",
       "      <td>0.002039</td>\n",
       "      <td>107.145</td>\n",
       "      <td>499784</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.244598</td>\n",
       "      <td>0.002434</td>\n",
       "      <td>100.512</td>\n",
       "      <td>339441</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Sicilian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.212272</td>\n",
       "      <td>0.001975</td>\n",
       "      <td>107.486</td>\n",
       "      <td>505477</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.231141</td>\n",
       "      <td>0.002260</td>\n",
       "      <td>102.297</td>\n",
       "      <td>351028</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Spanish_North</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.215885</td>\n",
       "      <td>0.002029</td>\n",
       "      <td>106.383</td>\n",
       "      <td>500853</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.238479</td>\n",
       "      <td>0.002426</td>\n",
       "      <td>98.319</td>\n",
       "      <td>341661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Spanish</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.213869</td>\n",
       "      <td>0.001975</td>\n",
       "      <td>108.297</td>\n",
       "      <td>513648</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.235386</td>\n",
       "      <td>0.002257</td>\n",
       "      <td>104.293</td>\n",
       "      <td>361951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Ukrainian</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.218716</td>\n",
       "      <td>0.002007</td>\n",
       "      <td>108.950</td>\n",
       "      <td>503981</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.243551</td>\n",
       "      <td>0.002345</td>\n",
       "      <td>103.881</td>\n",
       "      <td>348948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Levanluhta</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.236252</td>\n",
       "      <td>0.002383</td>\n",
       "      <td>99.123</td>\n",
       "      <td>263049</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.247640</td>\n",
       "      <td>0.003030</td>\n",
       "      <td>81.728</td>\n",
       "      <td>174148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>BolshoyOleniOstrov</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.247814</td>\n",
       "      <td>0.002177</td>\n",
       "      <td>113.849</td>\n",
       "      <td>457102</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.256041</td>\n",
       "      <td>0.002624</td>\n",
       "      <td>97.561</td>\n",
       "      <td>305851</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>ChalmnyVarre</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.233499</td>\n",
       "      <td>0.002304</td>\n",
       "      <td>101.345</td>\n",
       "      <td>366220</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.249619</td>\n",
       "      <td>0.002862</td>\n",
       "      <td>87.212</td>\n",
       "      <td>239594</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>result:</td>\n",
       "      <td>Han</td>\n",
       "      <td>Saami.DG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.236198</td>\n",
       "      <td>0.002274</td>\n",
       "      <td>103.852</td>\n",
       "      <td>489038</td>\n",
       "      <td>result:</td>\n",
       "      <td>MA1_HG.SG</td>\n",
       "      <td>Mbuti</td>\n",
       "      <td>0.251530</td>\n",
       "      <td>0.002622</td>\n",
       "      <td>95.922</td>\n",
       "      <td>326072</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   dummy_Han A_Han                   B  C_Han    F3_Han  stderr_Han    Z_Han  \\\n",
       "0    result:   Han             Chuvash  Mbuti  0.233652    0.002072  112.782   \n",
       "1    result:   Han            Albanian  Mbuti  0.215629    0.002029  106.291   \n",
       "2    result:   Han            Armenian  Mbuti  0.213724    0.001963  108.882   \n",
       "3    result:   Han           Bulgarian  Mbuti  0.216193    0.001979  109.266   \n",
       "4    result:   Han               Czech  Mbuti  0.218060    0.002002  108.939   \n",
       "5    result:   Han               Druze  Mbuti  0.209551    0.001919  109.205   \n",
       "6    result:   Han             English  Mbuti  0.216959    0.001973  109.954   \n",
       "7    result:   Han            Estonian  Mbuti  0.220730    0.002019  109.332   \n",
       "8    result:   Han             Finnish  Mbuti  0.223447    0.002044  109.345   \n",
       "9    result:   Han              French  Mbuti  0.216623    0.001969  110.012   \n",
       "10   result:   Han            Georgian  Mbuti  0.214295    0.001935  110.721   \n",
       "11   result:   Han               Greek  Mbuti  0.215203    0.001984  108.465   \n",
       "12   result:   Han           Hungarian  Mbuti  0.217894    0.001999  109.004   \n",
       "13   result:   Han           Icelandic  Mbuti  0.218683    0.002015  108.553   \n",
       "14   result:   Han       Italian_North  Mbuti  0.215332    0.001978  108.854   \n",
       "15   result:   Han       Italian_South  Mbuti  0.211787    0.002271   93.265   \n",
       "16   result:   Han          Lithuanian  Mbuti  0.219615    0.002032  108.098   \n",
       "17   result:   Han             Maltese  Mbuti  0.210359    0.001956  107.542   \n",
       "18   result:   Han           Mordovian  Mbuti  0.223469    0.002008  111.296   \n",
       "19   result:   Han           Norwegian  Mbuti  0.218873    0.002023  108.197   \n",
       "20   result:   Han            Orcadian  Mbuti  0.217773    0.002014  108.115   \n",
       "21   result:   Han             Russian  Mbuti  0.223993    0.001995  112.274   \n",
       "22   result:   Han           Sardinian  Mbuti  0.213230    0.001980  107.711   \n",
       "23   result:   Han            Scottish  Mbuti  0.218489    0.002039  107.145   \n",
       "24   result:   Han            Sicilian  Mbuti  0.212272    0.001975  107.486   \n",
       "25   result:   Han       Spanish_North  Mbuti  0.215885    0.002029  106.383   \n",
       "26   result:   Han             Spanish  Mbuti  0.213869    0.001975  108.297   \n",
       "27   result:   Han           Ukrainian  Mbuti  0.218716    0.002007  108.950   \n",
       "28   result:   Han          Levanluhta  Mbuti  0.236252    0.002383   99.123   \n",
       "29   result:   Han  BolshoyOleniOstrov  Mbuti  0.247814    0.002177  113.849   \n",
       "30   result:   Han        ChalmnyVarre  Mbuti  0.233499    0.002304  101.345   \n",
       "31   result:   Han            Saami.DG  Mbuti  0.236198    0.002274  103.852   \n",
       "\n",
       "    nSNPs_Han dummy_MA1      A_MA1  C_MA1    F3_MA1  stderr_MA1    Z_MA1  \\\n",
       "0      502678   result:  MA1_HG.SG  Mbuti  0.243818    0.002349  103.781   \n",
       "1      501734   result:  MA1_HG.SG  Mbuti  0.236494    0.002296  103.008   \n",
       "2      504370   result:  MA1_HG.SG  Mbuti  0.231399    0.002264  102.229   \n",
       "3      504310   result:  MA1_HG.SG  Mbuti  0.237498    0.002281  104.103   \n",
       "4      504089   result:  MA1_HG.SG  Mbuti  0.243224    0.002328  104.457   \n",
       "5      510853   result:  MA1_HG.SG  Mbuti  0.226740    0.002197  103.193   \n",
       "6      504161   result:  MA1_HG.SG  Mbuti  0.243135    0.002317  104.941   \n",
       "7      503503   result:  MA1_HG.SG  Mbuti  0.247065    0.002362  104.619   \n",
       "8      502217   result:  MA1_HG.SG  Mbuti  0.245684    0.002379  103.266   \n",
       "9      509613   result:  MA1_HG.SG  Mbuti  0.240235    0.002269  105.886   \n",
       "10     503598   result:  MA1_HG.SG  Mbuti  0.232645    0.002253  103.243   \n",
       "11     507475   result:  MA1_HG.SG  Mbuti  0.236566    0.002280  103.757   \n",
       "12     507409   result:  MA1_HG.SG  Mbuti  0.241720    0.002313  104.483   \n",
       "13     504655   result:  MA1_HG.SG  Mbuti  0.244488    0.002386  102.481   \n",
       "14     507589   result:  MA1_HG.SG  Mbuti  0.236407    0.002273  104.002   \n",
       "15     492400   result:  MA1_HG.SG  Mbuti  0.230839    0.002767   83.427   \n",
       "16     503681   result:  MA1_HG.SG  Mbuti  0.246864    0.002403  102.718   \n",
       "17     503985   result:  MA1_HG.SG  Mbuti  0.230200    0.002259  101.903   \n",
       "18     503441   result:  MA1_HG.SG  Mbuti  0.245284    0.002346  104.571   \n",
       "19     504621   result:  MA1_HG.SG  Mbuti  0.243930    0.002301  106.031   \n",
       "20     504993   result:  MA1_HG.SG  Mbuti  0.243614    0.002320  105.008   \n",
       "21     506525   result:  MA1_HG.SG  Mbuti  0.245212    0.002298  106.698   \n",
       "22     508413   result:  MA1_HG.SG  Mbuti  0.231967    0.002264  102.449   \n",
       "23     499784   result:  MA1_HG.SG  Mbuti  0.244598    0.002434  100.512   \n",
       "24     505477   result:  MA1_HG.SG  Mbuti  0.231141    0.002260  102.297   \n",
       "25     500853   result:  MA1_HG.SG  Mbuti  0.238479    0.002426   98.319   \n",
       "26     513648   result:  MA1_HG.SG  Mbuti  0.235386    0.002257  104.293   \n",
       "27     503981   result:  MA1_HG.SG  Mbuti  0.243551    0.002345  103.881   \n",
       "28     263049   result:  MA1_HG.SG  Mbuti  0.247640    0.003030   81.728   \n",
       "29     457102   result:  MA1_HG.SG  Mbuti  0.256041    0.002624   97.561   \n",
       "30     366220   result:  MA1_HG.SG  Mbuti  0.249619    0.002862   87.212   \n",
       "31     489038   result:  MA1_HG.SG  Mbuti  0.251530    0.002622   95.922   \n",
       "\n",
       "    nSNPs_MA1  \n",
       "0      350484  \n",
       "1      344332  \n",
       "2      349612  \n",
       "3      349800  \n",
       "4      349553  \n",
       "5      359004  \n",
       "6      349321  \n",
       "7      348861  \n",
       "8      347208  \n",
       "9      357842  \n",
       "10     349082  \n",
       "11     355261  \n",
       "12     355340  \n",
       "13     350287  \n",
       "14     354999  \n",
       "15     321217  \n",
       "16     348656  \n",
       "17     347725  \n",
       "18     350058  \n",
       "19     350182  \n",
       "20     351053  \n",
       "21     355953  \n",
       "22     355548  \n",
       "23     339441  \n",
       "24     351028  \n",
       "25     341661  \n",
       "26     361951  \n",
       "27     348948  \n",
       "28     174148  \n",
       "29     305851  \n",
       "30     239594  \n",
       "31     326072  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "outgroupf3dat_merged"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we can make a scatter plot:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 10))\n",
    "plt.scatter(x=outgroupf3dat_merged[\"F3_Han\"], y=outgroupf3dat_merged[\"F3_MA1\"])\n",
    "plt.xlabel(\"F3(Test, Han; Mbuti)\");\n",
    "plt.ylabel(\"F3(Test, MA1; Mbuti)\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This isn’t very useful, however, as we cannot see which point is which population. We can use the annotation function from matplotlib to add text labels to each point:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 10))\n",
    "plt.scatter(x=outgroupf3dat_merged[\"F3_Han\"], y=outgroupf3dat_merged[\"F3_MA1\"])\n",
    "for i, row in outgroupf3dat_merged.iterrows():\n",
    "    plt.annotate(row[\"B\"], (row[\"F3_Han\"], row[\"F3_MA1\"]))\n",
    "plt.xlabel(\"F3(Test, Han; Mbuti)\");\n",
    "plt.ylabel(\"F3(Test, MA1; Mbuti)\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The result shows that indeed the affinity to East Asians in the bulk of European contries can be explained by MA1-related ancestry. Most European countries have a linear relationship between their affinity to Han and their affinity to MA1. However, this is not true for our ancient samples from Fennoscandia and for modern Saami and Chuvash, who have extra affinity to Han not explained by MA1 ([Lazaridis et al. 2014](https://www.nature.com/articles/nature13673)).\n",
    "\n",
    "Now, why there is a connection between MA1 and Han is not trivial to explain. The most probable explanation involves \"Basal Eurasian\" ancestry, which happens to be anti-correlated to MA1-ancestry in Europe, and which drives those populations with high \"Basal Eurasian\" ancestry further away from Han. See [Lazaridis et al. 2014](https://www.nature.com/articles/nature13673) for more details."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}