454 Output Data Analysis

[jspall]

Hi All - I have just received the output files from 13-lanes of a 454 run and I have no idea what to do with them. I have 13 .fna files and 13 .qual files and don't know where to start.

The goal is to identify the sequences to species level. Does anyone know what I should do or where I could go for help?

thanks,

Jenn

[nickloman]

Seriously? Why did you do the sequencing then?!?

First you need to know if you've sequenced 16S amplicons or shot-gun metagenomics. If the former, I would suggest checking out QIIME. If the latter I'd look at MEGAN or MG-RAST.

BTW I doubt you have 13 lanes of sequencing, I would expect you have sequence from a single region which has been demultiplexed by barcode to 13 files.

[Mathgon]

You should find money for a postdoc... here I am :P

[jiaco]

Sadly, you miss the sff file. You need to read this:

http://www.mothur.org/wiki/Schloss_SOP

Mothur and Qiime can both process this type of data, but I find the Mothur wiki a bit more informative than the Qiime docs. Try them both.

[jspall]

Seriously? Why did you do the sequencing then?!?

First you need to know if you've sequenced 16S amplicons or shot-gun metagenomics. If the former, I would suggest checking out QIIME. If the latter I'd look at MEGAN or MG-RAST.

BTW I doubt you have 13 lanes of sequencing, I would expect you have sequence from a single region which has been demultiplexed by barcode to 13 files.

I guarantee you that I there were 13 lanes - I did the 454 prep myself and I am quite sure I loaded 16 lanes, 13 with my samples, 3 with a lab mate's samples. I sequenced three gene regions (16S, rbcL and ITS) per sample and one sample per lane for 13 samples. If my math is correct, which I'm quite certain it is, then 13 samples x one lane per sample = 13 lanes.

As for your original (rather rude and inane) question, I sequenced them because in order to do sequence analysis, I need sequences.

Thank you for your suggestion of QIIME, I am familiar with it as the developer is a close friend. Perhaps I would receive a more pragmatic response if I contact him directly.

[jspall]

You should find money for a postdoc... here I am :P

lol - I wish. Do you work for free?

[jspall]

Sadly, you miss the sff file. You need to read this:

http://www.mothur.org/wiki/Schloss_SOP

Mothur and Qiime can both process this type of data, but I find the Mothur wiki a bit more informative than the Qiime docs. Try them both.

Thanks! I'm not familiar with MOTHUR, I'll take a look.

[nickloman]

As for your original (rather rude and inane) question, I sequenced them because in order to do sequence analysis, I need sequences.

My question wasn't meant to be rude, I was expressing my surprise that you would do the sequencing first before you had an idea about how you would analyse the data. And if you knew about QIIME, why didn't you try that out first? They have great tutorials on their site.

BTW you have regions, not lanes. 454 uses a rubber gasket to subdivide the PTP, sounds like you used one that divided it into 16 regions. The Illumina instrument has physical lanes on the flowchip.

Here's a great guide for getting the best out of resources like Seqanswers.com in future:

http://www.ploscompbiol.org/article/info%3Adoi%2F10.1371%2Fjournal.pcbi.1002202