If I Have Ion Torrent Whole Transcriptome Data, How Can I Look For Specific Genes' Expression?

Question

[jabeller]

I am brand spanking new to this, recieved a pilot grant to use the Ion Torrent sequencing system. I recieved two .sff files (one per condition) that are 2 GB in size and I have no idea what to do with them. I'm interested in a set of specific genes and whether their expression changes by condition. Where do I start, what program do I need? Any help would be greatly appreciated!

To add more detail: I had extracted RNA from two cultures of the same cell grown in two different conditions. I had their transcriptome sequenced by an Ion Torrent machine. I want to look at the expression of a set of specific genes in the transcriptome of these cells. This is a pilot project and is not meant be conclusive, but more of a proof of principle.

I will be seeking assistance next week from the facility that ran my samples, just thought I could try to get a head start over the weekend. Cheers!

[Istvan Albert]

I would recommend to find a collaborator with experience in the field

[Michael]

I think that's the best you can do, however I could question what the reason might be not to give such grants to experienced researchers in the first place (but I don't!).

Answer 1

[Michael]

If you cannot find an experienced lab to collaborate with:

Your second best bet might be to find out as much about the protocol for data generation and then try to apply and to follow a (more or less) standardized protocol for RNA-seq analysis and to vary parameters as you go along:

e.g. look at this question http://www.biostars.org/post/show/6615/rna-seq-data-analysis-good-reviewresearch-article-on-same/#6646 Here is another overview: http://seqanswers.com/wiki/How-to/RNASeq_analysis

It is advisable that you read through several review papers and then reproduce an established analysis. The only difference is that you have to convert your sff files to FASTQ format to be usable with most software, then you can look at read quality and read lengths (might be longer than 454 reads, so you cannot really use short-read alignment tools either) and choose alignment tools according to this. The conversion of sff to fastq should now work with sffinfo. You should have access to it from the manufacturer, it is contained in the Newbler assembler suit too, and it could work on your files too.

As you only have two files, it looks like you have no replication (or you could have barcoded replicates in the files, unlikely, but that's why you need to know how your data was generated). Then you will be unable to assess biological/technical variation, and you will only be able to perform a light version of DE-analysis without very meaningful p-values, even though edgeR allows for computation of p-values without replication.

Then you can come back with more specific questions regarding individual steps of your analysis. Also it would be great if you posted your findings here. It is hard to give advice for analyzing data nobody has access to.

Answer 2

[jabeller]

The point of the pilot grant opportunity was to expand the use of the machinery since it is a campus facility, also I think my scientific question is an interesting and novel one.

Is it really that complex of an analysis that I wouldn't be able to figure it out on my own (I am tech-savvy)? I know I can look up the sequence for the mRNA using NCBI, but I don't know what program is best to open up the .sff file in the first place.

[Istvan Albert]

In response to Is it really that complex of analysis - as a scientists I am sure you recognize that it all depends on what you are trying to do. I would not underestimate the difficulty of extracting information from data.

[Michael]

please use comments instead of answers to comment

[Istvan Albert]

you may also click the edit link on your question and add more information

[jabeller]

I'm not underestimating, I'd just like to learn the technique of parsing through the data on my own. Sorry about posting that as an answer... new around here.

[caseyr547]

you want sfffile from 454 to open your sffs http://454.com/contact-us/software-request.asp if you need more technical help you can contact the company that i work for [e-mail]