Skip to content

Latest commit

 

History

History
127 lines (92 loc) · 4.34 KB

pong Guide.md

File metadata and controls

127 lines (92 loc) · 4.34 KB
Raw

pong Tutorial

By Pavel Salazar-Fernandez (epsalazarf@gmail.com)

Human Population and Evolutionary Genomics Lab | LANGEBIO

About

Documentation: pong (GitHub)

This document explains how to generate multiple ADMIXTURE data for integration in pong.

pong

pong is a package consisting of a new algorithm for fast post-hoc analysis of clustering inference output from population genetic data and a custom, interactive D3.jsbased data visualization. Our package solves the challenges of accounting for label switching, characterizing multiple modes, and aligning Q matrices across values of K by constructing weighted bipartite graphs for each pair of Q matrices depicting similarity in membership coefficients between clusters.

For installation instructions, an introductory tutorial and the detailed manual, check the software web page:

https://github.com/abehr/pong

Procedure

Requires:

  1. Get the PLINK binary files (.bed, .bim, .fam) to be analyzed. The dataset must contain only the samples of interest. Preferably, replace the samples' FAMIDs with their population tags in the .fam file for easier grouping and sorting.

  2. Run the multiple ADMIXTURE analysis with the following command (script is annexed at the end of this document): multiAdmix.sh [DATASET] [minimum K] [maximum K] [iterations] Be warned that depending on the number of samples, K's and iterations, this analysis can become very long.

  3. When all the analyses finishes, the results will be saved at a directory called [DATASET]-multiAdmix/, containing multiple run[##]/ directories filled with Q files at each specified K.

  4. Generate the following files required by pong at the [DATASET]-multiAdmix/ directory:

    • pong_filemap.txt: a list containing the location for all the Q files to be integrated by pong. The pongmap.sh can generate it automatically by running it in the "[DATASET]-multiAdmix/" directory (script included in the annex).
    • ind2pop.txt: a list with repeated populations tags used to replace the individual tags; must contain the same number of lines as samples. If you replaced the FAMID with populations tags, just run cut -f 1 [DATASET].fam > ind2pop.txt.
    • pop_order.txt: used to sort the populations in the display.
    • pop_order_expanded.txt: same as pop_order.txt, with a second column for extended names to replace the population tags in the plots.

  5. Run pong in the [DATASET]-multiAdmix/ directory with the following command: pong -m pong_filemap -n pop_order_expandednames.txt -i ind2pop.txt

  6. Once finished, a port for visualizing the results will start. Open a web browser (not IE) and go to the address http://localhost:4000.

Other functions

TBD

Scripts

multiAdmix.sh

#!/bin/bash

#MULTIADMIX.sh
# $1: Dataset name
# $2: Min K
# $3: Max K
# $4: Iterations

echo "Starting MultiAdmix..."

if [[ $2 = [0-9]* && $3 = [0-9]* && $4 = [0-9]* ]] && (($2 <= $3))

then
  echo "Dataset: " $1
  echo "Min K: " $2
  echo "Max K: " $3
  echo "Iterations: " $4

  mkdir $1-multiAdmix
  cd $1-multiAdmix
  for ((i=1; i<=$4; i++))
  do
    echo "Run " $i
    mkdir run$i
    x=$RANDOM
    cd run$i
    for ((k=$2; k<=$3; k++))
    do
      admixture --cv -s $x ../../${1}.bed $k > ${1}.r${i}k${k}.log
    done
    cd ..
  done

  echo "All runs finished!"
else
  echo "ERROR: non-integer in input parameters"
  echo "Usage: multiAdmix.sh {dataset} {minK} {maxK} {iterations}"
fi

#END

pongmap.sh

#!/bin/bash

runs=$(find -name "run*" | sed 's/\.\/run\([0-9][0-9]*\)/\1/' | sort -n | tail -n 1)
mink=$(find -name "*.Q" | rev | cut -f 2 -d "." | sort  -n | uniq | head -n 1)
maxk=$(find -name "*.Q" | rev | cut -f 2 -d "." | sort  -n | uniq | tail -n 1)
root=$(find -name "*.Q" | rev | cut -f 1  -d "/" | cut -f 3- -d "." | rev | head -n 1)

echo "Files name: " $root
echo "Runs found: " $runs
echo "MinK found: " $mink
echo "MaxK found: " $maxk

echo "Starting listing..."

touch pong_filemap.txt

for ((k=$mink; k<=$maxk; k++))
do
  echo "K:" $k
  for ((r=1; r<=$runs; r++))
  do
    echo -e "k${k}r${r}\t$k\trun${r}/${root}.${k}.Q" >> pong_filemap.txt
  done
done

echo "List finished! > pong_filemap.txt"

#END