Second,
Is any one on the mailing-list have Illumina-capable versions of
TopHat/Cufflinks, ones that don't require grooming ?
This would be pretty easy to do:
(1) change the input datasets to be fastq rather than fastqsanger in tophat_wrapper.xml;
(2) add an input to the wrapper to specify the input type, solexa 1.0, solexa/illumina
1.3+, sanger;
(3) modify tophat_wrapper.py to accept and process the new input type.
FYI, I think Tophat assumes sanger format by default.
J.