popPhyl_PCA

Performs PCA of genotypes.
Works in two steps.

1. Input file

A single fasta file containing different loci, in different populations/species. Not necessarily sorted.
The ID (the line starting by >) of each sequence has to respect the following format:
`

E24_99631_p1|arabidopsis|E15|Allele_1 NNNNNNNNNNNAAAGAAGATGGCGTCGGCAGTTTCAGTATCGTTTATTGTGGTGAATATT TTGCTTCTCCTGGTTCAGGTCTTTGCTGGGAGAGACTTTTACAAAATATTGGGAGTTCCC AGAAACGCCGATTTGAAACAAATCAAGCGATCCTATCGAAAGCTGGCCAAAGAACTCCAC CCAGATAAGAACAAAGATGATCCTGAAGCAGAACAAAGATTTCAAGACTTAGGTGCTGCT ` Four different fields separated by a pipe (|), where:

  1. first field is the locus name (E24_99631_p1).
  2. second field is the species name (arabidopsis).
  3. third field is the name of the sampled diploid individual (E15).
  4. fourth field is the name of the allele (two alleles per individual, named either Allele_1 or Allele_2)

1. PCA

Single python command line (popphyl2PCA.py).
Before, you need to have these python dependencies available:

  1. pandas
  2. sklearn
  3. biopython

python3 ~/Programmes/popPhyl_PCA/popphyl2PCA.py [name of the subdirectory created by the script where output files will be written] [name of the input fasta file]

Example:
python3 ~/Programmes/popPhyl_PCA/popphyl2PCA.py ~/Documents/PCA/testPCA ~/Programmes/popPhyl_PCA/test.fas
Can takes between 10 minutes and 2 hours, depending on the number of SNPs and individuals.

2. vizualisation

Little Shiny interface (plotPCA.R).
Before, you need to have these R dependencies available:

  1. shiny
  2. plotly
  3. tidyverse
  4. shinycssloaders

Then, in R:

  1. source(~/Programmes/popPhyl_PCA/plotPCA.R)
  2. shinyApp(ui=ui, server=server)
  3. upload the files with coordinates (table_coord_PCA_genotypes.txt) and eigen values (table_eigen_PCA_genotypes.txt)

GitHub

https://github.com/popgenomics/popPhyl_PCA