SRA from NCBI?

Question

[star]

I like to get some data to reanalysis them, so I start to get SRA from NCBI but I faced some question?

• for example: for SRR1067431 and SRR5331265, there are 5 runs with one GEO number and 1 run with a GEO number, respectively. Does it mean that there are 4 technical replicates for SRR1067431 that I can merge them from the first step (merge their FASTQ files)?

• is there any way to check md5sum for FASTQ file or I only should use vdb-validate from sratoolkits?

• why md5sum is changing, when FASTQ file change to FASTQ.GZ?

[GenoMax]

You will need to look through the metadata to see what the actual samples represent.

Additional download help: Fast download of FASTQ files from the European Nucleotide Archive (ENA) and https://ewels.github.io/sra-explorer/

Answer 1

[ATpoint]

The first one looks like a lane/sequencing replicate. As always with these, they probably can be merged right away into one file prior to alignment. Still, for quality control you could keep them separated and merge lateron. md5 is specific for every file, so of course (de)compression changes it. Any need to check that? Do you have indications that some are corrupted? I never felt the need to do that and never experienced any problems with sra files given the download completed properly.

[star]

Thanks, No they are downloaded completely.

[ATpoint]

Should be fine then. fastq-dump will throw an error anyway if something is corrupted so don't waste your time on md5 ;-)