Hi,

I have downloaded Illumina reads from SRA (SRR769347). I want to run a de novo assembly with SPAdes but I am getting an error and I am not sure how to fix it.

Here are all the steps I've performed:

I used fastq-dump --split-files on the SRA file to extract the 2 files for each paired reads. I used sed to add /1 and /2 respectively to each paired files. I therefore have 2 files: file1.fastq and file2.fastq

A read from file1.fastq looks like this:

@SRR769347.1/1 1 length=101
CCTGTGTGATAAAATTGGAAAACAAATTCAAACTGATACTGAAATCAAAGCTGCCGTGTTTGACTTAAAACAAATGCTTGATCACTAGCAGAAAAACCTGA
+SRR769347.1 1 length=101
@@@FFFDDAFDHHJIJIJGGIIIIGCHHIIGHIGIGCGHGIJFEE>GHHGEIJIEE@FFHIIGJIJCC@GGIFHHFEHFBDDDFECCC9>CDDD?AB@DB9

A read from file2.fastq looks like this:

@SRR769347.1/2 1 length=101
TGAATACCTTCTTTTTTAGCAAAAAATTGAATGTCATCCACAAGTAATAAGTCACAGGTGCGATACTTTTCTCTGAATTCATCCTGAGTTTTATTTTTGAT
+SRR769347.1 1 length=101
CC@FFDFFHHHHHIJJJJJIJIIIGIGHIIJJJJIIIIJEIGCHBFHIGGGGGGGGIG@FHGGGEHHFFFFFFFEAEEECCADDDC5;>CCAACCCDCDBA

I then use these files as input to SPAdes 3.1.0 (available version on our cluster) as follows : (I also provide PacBio reads for the assembly)

spades.py -1 file1.fastq -2 file2.fastq --pacbio pacbio.fastq -o SPAdes_output

Invariably I am getting the following error:

== Error ==  file not found: file2.fastq (right reads, library number: 1, library type: paired-end)

I also tried to use another fastq-dump option: --split3 for which the reads are correctly labelled as /1 and /2 but strangely each read is in multiple copies... and also gives me the same error in SPAdes...

Any help would be great!

thanks

Julien