8. Running Stats & Filtering of Bad Samples

We used the outputs from PaleoMix--v1.2.5 to create a summary file containing information on the mapping statistics of each sample. In addition, we used some scripts to create some heatmap plots to help in the identification of bad SAMPLES, and also create some auxiliary files based on these plots.

Gets statistics and creates an absence/presence heatmap:

xsbatch -c 30 --mem-per-cpu 13000 -J HeatMap --time 5-00 -- "$SCRIPTS/scripts/paleomix_summary2tsv.sh -t 30 -n 10 -k 300 -i ~/data/Pigeons/PBGP/PBGP--Analyses/Lists/PBGP--AllSamples--Article.labels ~/data/Pigeons/Analysis/PaleoMix_Re-Sequencing/ ~/data/Pigeons/Analysis/PaleoMix_GBS/ > ~/data/Pigeons/PBGP/PBGP--Analyses/PBGP--CoverageHeatMap/Stats_PBGP--Article--Ultra.txt"

These results were plotted using the Rscript below:

• PBG--CoverageHeatmap.R

Gets Cutsites Information

grep -v "WGS" ~/data/Pigeons/PBGP/FPGP--Analyses/PBGP--CoverageHeatMap/Loci_Merged.coverage.tsv | grep -v "Blank" | tail -n +2 | cut -f 2- | awk '{for(i=1; i<=NF; i++)x[i]+=$i} END{for(i in x)print x[i]}' > ~/data/Pigeons/PBGP/PBGP--Analyses/PBGP--CoverageHeatMap/Loci_Merged.coverage.cutsitesmath

awk '$1==0{cnt++} END{print cnt}' ~/data/Pigeons/PBGP/PBGP--Analyses/PBGP--CoverageHeatMap/Loci_Merged.coverage.cutsitesmath

Number of LOCI with No data for ALL: 288,319

We manually created a list containing SAMPLES to be excluded (6 BAD GBS SAMPLES and 2 BLANKS / highlighted on the Coverage HeatMap).

~/data/Pigeons/PBGP/PBGP--Analyses/Lists/PBGP--BadSamples--Article.list

Creates an ID file containing only scaffolds longer than 1Kb:

awk '$2 > 1000 {print $1":"}' ~/data/Pigeons/Reference/DanishTumbler_Dovetail_ReRun.fasta.fai > ~/data/Pigeons/Reference/DanishTumbler_Dovetail_ReRun_ChrGreater1kb.id

---