Rare NCBI SRA dataset - how can I use it?

[cebehind]

Hi to all,

I wanted to check this data and use cufflinks to get the FPKM values, but I haven't been able to. The SRA files are too light (200-300mb approximatly) and I know their usually around 1.5Gb.
I have little to none experience with processing this kind of data and I am having a hard time coming up with a solution for that. I will really appreciate if someone can explain me how to do that.

http://www.ncbi.nlm.nih.gov/geo/quer...i?acc=GSE68789 - This is the dataset.

Thanks a lot.

[cebehind]

It looks like the files you are referring to are files with the counts, not the raw data.

http://www.ncbi.nlm.nih.gov/sra?linkname=bioproject_sra_all&from_uid=283850

That's the link to SRA data, wich is supposed to be the raw files. But I'm not sure.

[mastal]

The first file listed looked like it was a summary of the counts.

The other files are .sra files, and I think they might be in a compressed format.

I think you need the sra toolkit to convert the files to something like fastq.

Have a look at:

http://www.ncbi.nlm.nih.gov/books/NBK158900/

[WhatsOEver]

http://www.ncbi.nlm.nih.gov/sra?link...rom_uid=283850

That's the link to SRA data, wich is supposed to be the raw files. But I'm not sure.

Yes, that is the link to the datafile in the SRA. There you can download the files and convert them to fastq via fastq-dump from the sra toolkit. After that you can go for mapping->cufflinks->whatsoever...

Alternatively, you can follow your first link to the geo deposit and download the counts table.

The read numbers are indeed not that high, but for DE it was probably enough - at least it was published with this data amount

[cebehind]

Yes, that is the link to the datafile in the SRA. There you can download the files and convert them to fastq via fastq-dump from the sra toolkit. After that you can go for mapping->cufflinks->whatsoever...

Alternatively, you can follow your first link to the geo deposit and download the counts table.

The read numbers are indeed not that high, but for DE it was probably enough - at least it was published with this data amount

http://www.ebi.ac.uk/ I used the European Bioinformatics Institute to download the fastq files. Would that work?

[mastal]

yes, that should be fine.

[cebehind]

I'm planing on using Galaxy then, I have no idea how to use R. Is there something that I must know beforehand?

[WhatsOEver]

The EBI hosts those as fastq? That's surprising but should be fine if you are sure the files are the correct ones. From the geo site you get some info on their analysis procedure which will be important if you want to (more or less) exactly reproduce the results (tophat --max-intron-length 10000 --max-multihits 1; Expression table done using ESAT http://garberlab.umassmed.edu/software/esat/; Genome_build: mm9).

[cebehind]

The EBI hosts those as fastq? That's surprising but should be fine if you are sure the files are the correct ones. From the geo site you get some info on their analysis procedure which will be important if you want to (more or less) exactly reproduce the results (tophat --max-intron-length 10000 --max-multihits 1; Expression table done using ESAT http://garberlab.umassmed.edu/software/esat/; Genome_build: mm9).

The thing is I wasn't able to run the ESAT I don't know why. So I'm aiming to get FPKM values for that. I'm having problems with the gene names in cufflinks. I don't know how to do that.