Help in sff file processing in amplicon analysis via mothur

[vinnie_assemble]

Hi,

I am using my 454 data for OTU analysis in mothur. And I am confused after transform my sff to a fasta file. Sequencing information, platform 454 FLX, flow pattern TACG, barcode (AAAAAAAC) removed by sequencing center, primer: GAGTTTGATCNTGGCTCAG.

However, I have trouble to understand the sequence section (from 5th base to 12nd base). The primer started from 13 base. I attached the output fasta format from different toolkit.

sff_extract (from seq_crumbs toolkit) with clipping:

GAGTTTGATCCTGGCTCAGATTGAACGCTGG....

sff_extract (from seq_crumbs toolkit) without clipping:

tcagagagcgaaGAGTTTGATCCTGGCTCAGATTGAACGCTGG...

mothur output after denoise:

AGAGCGAAGAGTTTGATCCTGGCTCAGATTGAACGCTGG...

Does anyone can help to understand the sequence agagcgaa part? Base on the sequencing center information, it does not belong to barcode. And how should I deal with it? For example, it there a way to remove this region in mothur? Thank you!

[martin2]

The 'agagcgaa' sequence is probably a custom MID (barcode). They (somebody at the sequencing centre) properly used the left trimpoint to delineate it. sff_extract does the right job as far as I can see.

I do not believe that anybody used 'AAAAAAAC' as a barcode as you say. That would be a good joke ... to design a homopolymer into a barcode for this platform (and also IonTorrent). Either way, the barcode would have to be visible in the sequence you showed but it is not. Nobody wrote a tool to edit the SFF files and slice them (only tools to 'mask' the existing sequence exist) so that is another reason why I do not believe somebody gave you SFF files with barcodes physically removed. Also, your sequence starts with the sequencing key 'tcag' so another reason to believe this is just the original, raw read sequence.

[vinnie_assemble]

Hello martin2,

Thank you for your help. I tried process_sff.py in QIIME, and they will trim the AGAGCGAA automatically. And I contact the sequencing center, they said AAAAAAAC are the artificial barcode, which being add after sequencing.

[martin2]

Thank you for your help. I tried process_sff.py in QIIME, and they will trim the AGAGCGAA automatically.

That is expected, as I said, that is very likely the real sample barcode. Provided it is clearly present in your reads I do not see any reason why should you believe there is yet another sample barcode elsewhere.

And I contact the sequencing center, they said AAAAAAAC are the artificial barcode, which being add after sequencing.

Do you see the AAAAAAAC sequence anywhere? I don't at least in those reads you posted.

In case you do not agree with the 'agagcgaa' idea go ahead and complain to your sequencing centre and tell them they gave you files with data of somebody else. ;-) Either way, if somebody picked up a AAAAAAAC as a sample barcode for 454 or IonTorrent technology then do not use their services anymore. It would be often destroyed by sequencing errors.