Genomescope and Jellyfish

Genomescope is a web-tool used to estimate genome size, heterozygosity, and repeat content from short read sequence data using a kmer-based statistical approach. To read more, or to access, visit Genomescope. To run Genomescope, you will need to apply Jellyfish, a tool used to quickly and efficiently count kmers in DNA. Apply Genomescope and Jellyfish in the following steps:

1) Count kmers using jellyfish

2) Generate histogram using jellyfish

3) Download histogram, and submit it to Genomescope

1 Count K-mers using Jellyfish

k-mer - all the possible nucleotide sequences of a certain length, ex. K=2 all the possible kmers are: AA AT AC AG TA TT TC TG CA CT CC CG GA GT GC GG Be sure the Jellyfish module is available and uploaded on your HPC.

First, unzip your files

#applying an * will unzip all .fastq.gz files
gzip -d *.fastq.gz

Then, write and submit a job script with the following code:

echo + `date` job $JOB_NAME started in $QUEUE with jobID=$JOB_ID on $HOSTNAME
#
jellyfish count -C -m 21 -t $NSLOTS -s 800000000 /path/to/fastq/files/*.fastq -o reads.jf 
#
echo = `date` job $JOB_NAME done
#
#Explanation
#-m = kmer length
#-s = RAM
#-C = "canonical kmers"

2 Generate a histogram file

Write and submit a job script with the following code:

echo + `date` job $JOB_NAME started in $QUEUE with jobID=$JOB_ID on $HOSTNAME
#
jellyfish histo -t $NSLOTS reads.jf > reads.histo
#
echo = `date` job $JOB_NAME done

3 Download histogram, and submit it to Genomescope

First, download the histogram file using ffsend

ffsend upload reads.histo

Next, upload the histogram file to Genomescope.

Record the url that Genomescope generates for you, so you can easily return to view your results in the future.