Suppose you do de novo assembly of a mammalian genome. How do you calculate the N50 value? I have tried quast but its fails probably because of the huge size of the genome.
Hi Ishtiaque,
you could try GNX-tools. It is a bit older, but still works great. See link below, installation instructions are in the git repository.
https://github.com/mh11/gnx-tools
You could then run the tool like this from the gnx folder (if you dont have it in your PATH):
java -jar gnx.jar ./PATH_TO_YOUR_GENOME.fasta
# Output will be like this:
Total number of sequences: 12345
Total length of sequences: 123,567,890 bp
Shortest sequence length : 2022 bp
Longest sequence length : 1,364,562 bp
Total number of Ns in sequences: 50
N50: 801234 (55 sequences) (152379250 bp combined)
-> You may have to increase your java memory limit, if you have got a very big genome. To something like this:
Picked up _JAVA_OPTIONS: -Xmx14096m
I guess that you have got it on a hpc somewhere that has already additional software installed, right? You can also talk to your hpc admin to check what software they have got installed already.
You could give this tool a try as well "assembly-stats":
https://github.com/sanger-pathogens/assembly-stats
Hi Ishtiaque
Did you try QUAST? It is a good tool, and I think it might work with large genomes also.
There is a web interface (http://cab.cc.spbu.ru/quast/) but it is limited to 250MB inputs, but you can install it locally.
Ricardo Franco-Duarte I have already tried QUAST but for some reason, it keeps failing in the local machine.
Ishtiaque Ahammad is I can be of any help, just write me an email: [email protected].
- Badges
- Science topic