Accurate estimating genome size is a crucial task in sequencing projects. Current methods often struggle with polyploidy or become inefficient when dealing with species that exceed a ploidy level of six. To address these challenges, we introduce findGSEP, an enhanced version of findGSE. findGSEP utilizes a segmented fitting approach to fit a normal distribution to polyploid species within a segmented framework. This ap-proach simplifies the process of single fitting while significantly expanding the range of ploidy levels it can handle. Moreover, findGSEP offers users interactive tools through both an open-source R application and a web application, facilitating reliable and precise estimation of genome size.

We have released our backend-server findGSEP and provide a CPU-based version of findGSEP online platform. Please check it out!!!

The workflow of the whole process consists of two main steps: Step 1. Generate histo file using either KMC or Jellyfish tools. Step 2. Run findGSEP.

1. Download and install jellyfish from: Jellyfish Release

2. Count kmers using jellyfish:

   jellyfish count -C -m 21 -t 1 -s 5G *.fastq -o reads.mer

   Note: Adjust the memory (-s) and threads (-t) parameters according to your server. This example uses 1 thread and 5GB of RAM. The kmer length (-m) may need to be scaled if you have low coverage or a high error rate. Always use 'canonical kmers' (-C).

3. Export the kmer count histogram:

   jellyfish histo -h 3000000 -t 10 -o reads.histo reads.mer

   Note: The thread count (-t) should be scaled according to your server.

4. Upload reads.histo to findGSEP.

1. Download and install KMC from: KMC GitHub

2. Count kmers using KMC:

   kmc -k21 -fa -t50 -m12 -ci1 @input_file_name.lst reads_kmc tmp

   *Note: input_file_name.lst is the file name list with locations. -fa means input file is fasta format data. Adjust the memory (-m) and threads (-t) parameters according to your server. This example uses 50 thread and 12GB of RAM as default. The kmer length (-k) may need to be scaled if you have low coverage or a high error rate. The -ci1 option ensures that kmers with a count of at least 1 are included.

   demo usage:

   kmc -k27 -m24 NA19238.fastq NA.res /data/kmc_tmp_dir/
   kmc -k27 -m24 @files.lst NA.res /data/kmc_tmp_dir/
   
   NA19238 - single file in specified (-f switch) format (gziped or not)
   
   @files - file name with list of input files in specified (-f switch) format (gziped or not)
   
   

   For details, please refer to KMC GitHub *

3. Export the kmer count histogram:

   kmc_tools transform reads_kmc histogram reads_kmc.histo

   Note: This will create the histogram file reads_kmc.histo.

4. Upload reads_kmc.histo to findGSEP or corresponding file location for running.

install.packages("findGSEP")

1. Install devtools:

install.packages("devtools")

2. Install directly from GitHub:

devtools::install_github("sperfu/findGSEP")

You can check our demo dataset at our webserver or drive for complete data. We have provide precalculated histo file whose ploidy number ranging from tetraploid to octoploid.

Taking a precalculated simulated Pentaploid data as an demo, named your_test_file.histo.

# Set options (optional):

options(warn = -1)

# Define input parameters:

path <- "histo_files"
samples <- "your_file.histo"
sizek <- 21
exp_hom <- 200
ploidy <- 4
output_dir <- "outfiles"
xlimit <- -1
ylimit <- -1
range_left <- exp_hom * 0.2
range_right <- exp_hom * 0.2

#Call the findGSEP function with specified parameters:

findGSEP(path, samples, sizek, exp_hom, ploidy, range_left, range_right, xlimit, ylimit, output_dir)

# For any questions, usage inquiries, or reporting potential bugs, please contact the author.

After running, You will find 'your_file.histo_hap_genome_size_est.pdf' in your output_dir folder, please give it a try!!!

You can reference to our paramenter setting for those species we used in our webserver or demo dataset.

If you enconter problem when installing devtools, especially for those packages below, please consider install them using conda install command:

conda install -c conda-forge r-gert
conda install -c conda-forge r-textshaping
conda install -c conda-forge r-ragg
conda install -c conda-forge r-pkgdown

If you enconter issues like:

1. could not find function "brewer.pal"

2. could not find function "alpha"

Solutions:

library(RColorBrewer)
library(ggplot2)