I am about to conduct my first metabarcoding study. We will be using the ITS amplicon to look at fungal community composition in soil samples.

I need to write a data management plan before I actually do this, and the guidelines state that the expected size of each dataset should be given (in the interests of arranging suitable storage strategies). I realise that the amount of data generated by metabarcoding probably varies a lot depending on things like sequencing depth (we are going with 100k tags per sample), nature of the samples, etc.

But I am clueless -- I have no idea even what order of magnitude we are talking here! Terabytes? It doesn't need to be particularly accurate for the data management plan. If anyone is able to suggest a rough, ballpark figure, I'd be very grateful. For instance, if anyone here has done metabarcoding themselves and could let me know what the file sizes were like that would suffice, barred they can tell me how many samples they had, so I can scale it to my own sample count.

FWIW, Google has been no help on this.

Edit in response to gringer: Sorry I did not think to mention this. We will probably be using a mixture of Illumina NovaSeq 6000 and Illumina MiSeq (two separate parts of the project).

Thanks for your help!

  • $\begingroup$ What sequencing platform will you be sequencing on? That will be the biggest factor in the amount of data generated for each sequencing run, as there's [usually] a fixed yield that would be divided by the number of samples. $\endgroup$
    – gringer
    Commented Apr 21 at 4:31

1 Answer 1


A single metagenomics barcoding file is small 10 Gigs max (unzipped). So simply multiply that by the number of samples you have, but per project realistically less than 50 samples is normal. Thus is 0.5 Tb is the maximum size. However, the file remains zipped at all times and downstream processing takes place from zipped file (it chunk loads and unzips it within the package), so the actually size consumed is alot smaller and depends on the zipping protocol. Most metagenomic packages fall into this category.

As long as you don't want to blast across 50 samples, which means unzipping each file in turn and would result in 0.5 Tb storage, you are good to go on minimal storage, which I'd estimate around 100 Gigs, with max. zipping, on a large single project (50 samples being a large project).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.