Java out of memory exception

[dvirdi01]

I ran bloom on multiple fastq files together by giving it 200GB ram. It gave me out of memory exception mid way. I was running the command using a salloc allocation on the server and for that I need to specify the amount of space I am allocating for it on the server. I gave it 200GB to begin with but I guess it wasn't enough. I there a way to see how much space it would need for all the jobs beforehand?

Also, I want to produce a transcriptome.fq file for each of my 17 input files. I put the path to all input files inside a txt file and gave that txt file as the input. Will it go through each file and create a transcriptome.fq file for each one separately (desired output) or would it create only one transcriptome.fq file at the end?

[kmnip]

Can you report your entire log? How many reads are there in total?

If you want to assemble 17 transcriptomes for each of 17 input FASTQ files individually, then you need to have 17 RNA-Bloom commands, one for each FASTQ file. The txt file input is for assembling multiple files in a single run.

[dvirdi01]

If I run it through snakemake by giving it all 17 input files would it again output just a single transcriptome.fq at the end? Or one for each input file?

[kmnip]

If you provide multiple FASTQ files to RNA-Bloom in one command, RNA-Bloom will create one assembly. E.g.

rnabloom -long reads_01.fastq.gz reads_02.fastq.gz reads_03.fastq.gz ...

If you have 3 separate RNA-Bloom commands, then there will be 3 assemblies. E.g.

rnabloom -long reads_01.fastq.gz -outdir outdir01 ...
rnabloom -long reads_02.fastq.gz -outdir outdir02 ...
rnabloom -long reads_03.fastq.gz -outdir outdir03 ...

Also, the final output assembly of RNA-Bloom is a FASTA file, not a FASTQ file, as mentioned in the README: https://github.com/bcgsc/RNA-Bloom#final-output-files-1