Alternative to cufflinks? - Bam to count table?

[hbt]

From the numerous posts here on seqanswers it's becoming increasingly evident that cufflinks and cuffdiff have some major issues relating to calculation of DE significance across multiple samples.

As such we are looking for suitable alternatives for DE calculation of tophat mapped datasets.

There area number of R based software packages available (like edgeR, DEseq, DEGseq) but all these rely on the data being in a count table for input into R.

Can anyone suggest the best/easiest way of transforming the data in .bam files to count tables?

I believe this can be done with HTSeq but are any of you using alternative methods that you can recommend?

Thanks in advance

[turnersd]

I've only used HTSeq. I have heard talk about eXpress, but haven't seen a publication yet.

[mgogol]

bedtools can do it (coverageBed). We typically count reads in uxons of a gene (unioned exon space). We're in the same boat with cufflinks - exploring alternatives...

[syfo]

An alternative to Cufflink/diff and to read count methods is the Flux Capacitor