Continuing with the admixture analysis with our new reference 3 dataset.
Here's the results spreadsheet for K=6.
You can click on the legend to the right of the bar chart to sort by different ancestral components.
Fst divergences between estimated populations for K=6 in the form of MDS plots.
And the numbers:
C1 C2 C3 C4 C5
C2 0.088
C3 0.085 0.132
C4 0.097 0.145 0.067
C5 0.165 0.203 0.182 0.171
C6 0.154 0.128 0.176 0.205 0.269
UPDATE: With fixed Reference 3.
Continuing with the admixture analysis with our new reference 3 dataset.
Here's the results spreadsheet for K=5.
You can click on the legend to the right of the bar chart to sort by different ancestral components.
Fst divergences between estimated populations for K=5 in the form of MDS plots.
And the numbers:
C1 C2 C3 C4
C2 0.078
C3 0.088 0.126
C4 0.153 0.176 0.127
C5 0.160 0.165 0.201 0.266
UPDATE: With fixed Reference 3.
Continuing with the admixture analysis with our new reference 3 dataset.
Here's the results spreadsheet for K=4.
You can click on the legend to the right of the bar chart to sort by different ancestral components.
I have implemented something in this bar chart. When you click on the legend to sort by a specific component, it filters the results so that only those populations with at least 5% of that component are shown. Let me know what you think about this.
Fst divergences between estimated populations for K=4 in the form of an MDS plot.
And the numbers:
C1 C2 C3
C2 0.120
C3 0.158 0.198
C4 0.168 0.124 0.267
UPDATE: With fixed Reference 3.
Let's start with some admixture analysis with our new reference 3 dataset. We already did K=2 admixture run.
Here's the results spreadsheet for K=3.
You can click on the legend to the right of the bar chart to sort by different ancestral components.
Fst divergences between estimated populations for K=3 in the form of an MDS plot.
And the numbers:
C1 C2
C2 0.114
C3 0.157 0.197
UPDATE: With fixed Reference 3.
Let's start with some admixture analysis with our new reference 3 dataset.
Here's the results spreadsheet for K=2 (i.e. two ancestral populations).
You can click on the legend to the right of the bar chart to sort by different ancestral components.
Fst divergences between estimated populations for K=2:
C1
C2 0.168
With the increase in the number of groups to more than 160, the bar chart has gotten too busy and is hard to figure out. Any suggestions to improve its readability? I have two ideas: One is to have a dropdown menu to select regions and thus have separate charts for each region. The other, which I like better, is to show only the top 50 groups for the ancestral component you are sorting by. Any other ideas are welcome.
I have fixed the problem with Reference 3 but if you notice any strange results, do let me know.
While the Reference 3 admixture results were generally good (and I have some nice surprises on the way I hope), the Reich et al populations had some weird behavior. From one K value to the next, their admixture would swing wildly especially among the minor components.
For example, for Chenchu, the 2nd component after South Asian was Southwest Asian (42%) at K=6, European (45%) at K=7 and American (32%) at K=8. That just didn't make any sense. It was similar for other Reich et al populations, but all the other reference populations seemed pretty stable.
The issue was that when I was creating Reference 3, I had to juggle lists of SNPs to figure out a way to include Reich et al with a large (>100,000) number of SNPs in the dataset since Reich doesn't have as many SNPs in common with the other datasets plus 23andme (v2 and v3) and FTDNA. In that effort where I was doing lots of SNP set intersections and unions I messed up. I used 217,000 SNPs. While these SNPs were present in all the other datasets, Reich et al had only 102,000 SNPs common with that set. Ouch! This was a royal mess as the high missing rate of Reich et al caused weird instability in its admixture results even though the rest of the results were mostly stable.
Now, I have pared down Reference 3 to 118,000 SNPs. These have a low missing rate in all the datasets. So I don't expect the same problems.
I am redoing the admixture runs with this new data and will have some of the results up soon.
Behar as in the Behar et al paper/dataset and not the Indian state of Bihar. The Behar dataset contains 4 samples of Paniya, which apparently is a Dravidian language of some Scheduled Tribes in Kerala.
I had always been suspicious of those four samples since one of them had admixture proportions similar to other South Indians but the other three were like Southeast Asians.
When I got the Austroasiatic dataset, I found out that they had the four Paniyas from Behar et al in their data. However, only one of those four was the same as Behar. The other three were different. So I now had 7 Paniya samples.
Let's look at the K=12 admixture results for these Paniyas.
Behar's GSM536916 was the one which was the same as Austroasiatic's D36 and it has regular South Indian results. The other three Behar Paniyas are very Southeast Asian (yellow in the plot) while the three Paniyas from Austroasiatic data are similar to GSM536916/D36.
Since the Austroasiatic Paniya samples originated from Behar et al, I guess at some point before the Behar data being submitted to the GEO database the Paniyas got mislabeled.
I am now excluding the four Paniyas from Behar et al dataset and only using the Paniya samples from Austroasiatic dataset.
Even though the Pan-Asian dataset is not public, there was a request for my script to convert the data to Plink's PED format.
Here is how I convert the Pan-Asian data to Plink's transposed file format.
#!/usr/bin/perl -w $file="Genotypes_All.txt"; open(INFILE,"<",$file); open(TFAM,">","panasian.tfam"); open(TPED,">","panasian.tped"); $line = <INFILE>; chomp $line; @first = split('\t',$line); foreach my $sample (5..$#first) { print TFAM "0 $first[$sample] 0 0 0 -9\n"; } my $alleles; while(<INFILE>) { chomp; @lines = split('\t',$_); my ($major,$minor) = split('/',$lines[4]); print TPED "$lines[2] $lines[1] 0 $lines[3]"; foreach my $snp (5..$#lines) { if ($lines[$snp] == 0) { $alleles = "$major $major";} elsif ($lines[$snp] == 1) { $alleles = "$major $minor";} elsif ($lines[$snp] == 2) { $alleles = "$minor $minor";} else { $alleles = "0 0";} print TPED " $alleles"; } print TPED "\n"; } close(INFILE); close(TFAM); close(TPED); |
Again, no guarantees! It's Perl though, so it should be more stable across various operating systems.
I have withdrawn the Admixture results for Reference 3 for now while I figure out why a few of them were weird and unstable.Далматин
I will report back on what I find and will have fixed results soon.
Following mallu's request, here is the code I used to convert Xing et al data to Plink's PED format.
#!/bin/bash dos2unix *.csv head --lines=1 JHS_Genotype.csv > header.txt awk -F, '{for (i=2;i<=NF;i++) print "0",$i,"0","0","0", "0"}' header.txt > xing.tfam sed '1d' JHS_Genotype.csv > genotype.csv sort -t',' -k 1b,1 genotype.csv > genotype_sorted.csv sort -t',' -k 1b,1 JHS_SNP.csv > snp_sorted.csv join -t',' -j 1 snp_sorted.csv genotype_sorted.csv > xing_compound.csv awk -F, '{printf("%s %s 0 %s ",substr($2,4),$1,$3); for (i=6;i<=NF;i++) printf("%s %s ",substr($i,1,1),substr($i,2,1)); printf("\n")}' xing_compound.csv > xing.tped plink --tfile xing --out xing --make-bed --missing-genotype N --output-missing-genotype 0 |
I make no guarantees that it will work for you. I used it on my Ubuntu box, but I am sure it'll have trouble on Mac OS.
Recent Comments