Problems with fastq-dump --split-files returns only one file for paired-end sample

Question

[peanut]

I have downloaded SRR files (using prefetch). It says on the ncbi website that all of my runs have 2 reads per spot, but when I for example run SRR19749981 with: fastq-dump --split-files SRR19749981, I only get one single fastq output. I only get SRR19749981_1.fastq, when I am expecting SRR19749981_1.fastq and SRR19749981_2.fastq.

What am I doing wrong?

[GenoMax]

While NCBI seems to indicate that this is a paired-end dataset both NCBI and EBI are generating single files.

Under the original data format tab two files seems to be available from Amazon AWS. You could get those (payment may be required since egress is not free).

s3://sra-pub-src-3/SRR19749981/RYA_1.fq.gz.1    
s3://sra-pub-src-3/SRR19749981/RYA_2.fq.gz.1

Otherwise email SRA support and let them know that only one file is being generated when the data is supposed to be paired end.

[peanut]

I have never encountered this problem before, but with your guidance, I have had an epiphany and should be able to solve it. Thank you very much.