hello,

There are no .CEL and .CDF files for RNA-Seq. Such files can be found for Affymetrix (microarray) experiments.

If you want to download the raw data you have to download the .SRA files (SRA are compressed reads files)

for E-GEOD-67785, follow the link to GEO : http://www.ncbi.nlm.nih.gov/projects/geo/query/acc.cgi?acc=GSE67785

then, in the bottom of the web page, next to "SRP/SRP057/SRP057087", click on "ftp" you will reach the following link : ftp://ftp-trace.ncbi.nlm.nih.gov/sra/sra-instant/reads/ByStudy/sra/SRP%2FSRP057%2FSRP057087

If you only want the preprocessed data, you can download them by clicking on "http" next to GSE67785_RAW.tar ...... or on the arrayExpress page you can click on "E-GEOD-67785.processed.1.zip"

good luck

DAVID Please tell me how to analyze the RNA seq data and which online server or tool is good.

There are 2 diffrent ways to do that.

1) you have to download raw data as SRA files (compressed format)

- convert .SRA files into .fastq files using sratoolkit/2.5.1 (http://www.ncbi.nlm.nih.gov/books/NBK158900/#SRA_download.how_do_i_use_the_sra_toolki)

- then you can analyse the data (trimming, mapping, transcript quantification,...). For that, you can work on a local platform (on linux by using libraries such as TOPHAT, CUFFLINKS or others. There are other softwares that can be used, some of them can be used on Windows )... I use linux.

Or, you can use web-based platforms such as Galaxy ( https://usegalaxy.org/). Then upload the raw data and analyse them using on line tools.

2) You can only download preprocessed data. For example, for E-GEOD-67785, download the file GSE67785_RAW.tar here : http://www.ncbi.nlm.nih.gov/geo/download/?acc=GSE67785&format=file

Then, for each sample you will have a table with : Gene names, FPKM (Fragments Per Kilobase Of Exon Per Million Fragments Mapped) that reflects normalized expression of a given gene, and Count that reflects non-normalized expression of a given gene (i.e number of reads mapped on a given locus).

Yo can use FPKM or Counts for the analysis. I use diffrent R packages for that (e.g Limma).

The problem with the second alternative is that you can not control the preprocessing step (i.e reads filtering, mapping, ...)