What does cufflinks compare BAM file to, without reference?

[Stefano Manzini]

Hello everybody!

Main question: with a minimal syntax like this

cufflinks -p 3 target.bam

what does cufflinks use to assembly the information contained in target.bam file?

I'd also like to know whether using a local reference (the fat 16GB+ files you can download from cufflinks website) cufflinks will name the transcripts with the gene names instead of CUFF.65279 and the like.

I'm a wet biologist trying to learn some bioinformatics. I managed to use bwa and picard to process the raw Illumina reads and get to a .bam file, and now I'm trying to learn how get some DE data out of my samples, but I hadn't found any tutorial clear enough to explain what cufflinks is actually doing. This is also why my thoughts are hazy and my questions are -probably- not outstanding.

Thank you for yuor help and support, as well as anyone who wrote some beautiful tutorials on SEQwiki.

[Stefano Manzini]

Update: I have managed to find some aid and comfort in a nicely written tutorial. If you're a newbie (like me), you don't know very well what you are doing (like me) and you're still awed by the magic behind cufflink/cuffdiff, you should try and read these:

http://biowhat.ucsd.edu/homer/basicTutorial/

especially this:

http://biowhat.ucsd.edu/homer/basicTutorial/rnaseqCufflinks.html