Stacks - handling RadTag data

Here I will place all the commands required to run stacks from my vm

Demultiplexing:

First, I demultiplexed all the output files from the sequencher, using the provided barcodes. To do this, I use build_samples_5_1.sh and build_samples_5_2.sh

Check if any individuals sequenced on plate 5 were already sequenced in previous plates.

I then compare if plate5 from Oregon contains any previously sequenced individual. I did this with the following this script for both cross G and cross AA.

Building stacks tags

To generate all the necessary stacks tags, I ran from shell build_tags_Goo.sh and build_tags_AAo.sh

Saving the genotype for genotype-phenotype matrix manipulation.

[root@balt-vm4-dario mysql]# export_sql.pl -D NfGo_radtags -b 11 -a geno -f ./NfGo_allgeno.xls -o xls -m gen -c -F mark=Any -F snps_l=1 -F snps_u=1 -F alle_u=4
[root@balt-vm4-dario mysql]# export_sql.pl -D NfAAo_radtags -b 12 -a geno -f ./NfAAo_allgeno.xls -o xls -m gen -c -F mark=Any -F snps_l=1 -F snps_u=1 -F alle_u=4

These output files then are saved as csv files.
On 19-Nov-2013.py I start generating the phenotype-genotype matrix.

Converting phenotype-genotype data files into ped files

File worked on 22-Nov-2013, where I removed F1 genotypes with less than 25% coverage.
Next, I excluded the markers with less than 20% coverage with 24-Nov-2014.py
With 27-Nov-2014.py I split the ped file in family-ped files, and with 27-Nov-2013_2.py I developed a method to infer F1 genotype based on P0 and F2 genotypes, family by family. Importantly, this method won't be used for the genome-evolution-survival QTL paper. For completeness, 28-Nov-2013.py takes a ped file, splits it into families, and computes the F1 genotypes. It saves everything in the same folder.