Skip to content

findGSE is a tool for estimating size of (heterozygous diploid or homozygous) genomes by fitting k-mer frequencies iteratively with a skew normal distribution model.

Notifications You must be signed in to change notification settings

HeQSun/findGSE

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

48 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

findGSE

findGSE is a tool for estimating size of (heterozygous diploid or homozygous) genomes by fitting k-mer frequencies iteratively with a skew normal distribution model, which is written in R (code). The current version works on Linux & Mac OS X with R version 3.3.1 or above.

To use findGSE, one needs to input a k value and a corresponding k-mer histo file generated with short reads, which contains two tab-separated columns. The first column gives frequencies at which k-mers occur in reads, while the second column gives counts of such distinct k-mers (example).

Given multiple fastq.gz files, here is a two-step example for counting k-mers with jellyfish:

  zcat *.fastq.gz | jellyfish count /dev/fd/0 -C -o test_21mer -m 21 -t 1 -s 5G
  jellyfish histo -h 3000000 -o test_21mer.histo test_21mer

After getting the .histo file, supposing findGSE has been installed (INSTALL), we can do the following for GSE under R environment:

  library("findGSE")
  findGSE(histo="test_21mer.histo", sizek=21, outdir="hom_test_21mer")

Results will be printed like "Genome size estimate for test_21mer.histo: 1498918 bp." For more information about estimation, one can check the .txt and .pdf files in the output dir.

Two detailed toy examples about GSE for heterozygous and homozygous genomes are provided for playing around.

About

findGSE is a tool for estimating size of (heterozygous diploid or homozygous) genomes by fitting k-mer frequencies iteratively with a skew normal distribution model.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • R 78.5%
  • Perl 10.0%
  • C++ 8.5%
  • Shell 2.9%
  • Makefile 0.1%