md5sum got different hash code of SRA files

[emanlee]

Hi, All
We downloaded ftp://ftp-trace.ncbi.nlm.nih.gov/sra.../SRR308015.sra
and check its md5sum using the following command on linux:
md5sum SRR308015.sra
we got results:
5205d800cef6eff41beaa94f4a00b0e9
Which is different from hash code released by SraRunInfo.csv from SRA:
RunHash
0602E4F9BF0EDC10871B6AD6A4136CF4
ReadHash
D5D02A4031F17B6CA30ED7EBE52D5654

Thanks!

Aimin

[dariober]

Hi, All
We downloaded ftp://ftp-trace.ncbi.nlm.nih.gov/sra.../SRR308015.sra
and check its md5sum using the following command on linux:
md5sum SRR308015.sra
we got results:
5205d800cef6eff41beaa94f4a00b0e9
Which is different from hash code released by SraRunInfo.csv from SRA:
RunHash
0602E4F9BF0EDC10871B6AD6A4136CF4
ReadHash
D5D02A4031F17B6CA30ED7EBE52D5654

Thanks!

Aimin

I've downloaded the sra you linked and got md5sum:

5205d800cef6eff41beaa94f4a00b0e9

Are you sure the hashes (RunHash, ReadHash) you are looking at are md5sum?

[emanlee]

Thanks! We are not sure what hash code were used in Runhash and Readhash.
We have seen such a post:

http://www.ncbi.nlm.nih.gov/books/NBK47529/#_SRA_Quick_Sub_BK_Run_

[matheus.cburger]

Answer from NCBI support:

"That field you mentioned (RunHash) is an field used in internal file checking that has nothing to do with MD5 value. In fact SRA ftp files does not have that value. You will need to use the vdb-validate tool to validate a downloaded SRR#.sra file."