This tool has specific input requirements - namely, blast output. Usage details are on the tool form itself. A preview of this can be seen in the Tool Shed repo, here:
The Tuxedo suite does not produce consensus sequences, so using that output in a direct way will not work with this specific tool (meaning: extract transcript into a fasta file, run the blast, etc.). The coordinates can be extracted and formatted into BED format and used with the tool "Fetch Sequences -> Extract Genomic sequences" BUT be aware that this will be genomic sequence, none of the unique sequence content from the input samples will be included (base variations, etc). Still, that is an option.
One could also "annotate" the coordinates with either protein accessions for input to this tool, or directly with protein accessions with GO already assigned, using a method similar to the those described in this post. Overlapping coordinates on the same exact reference genome or some common key is all that is needed. It is very easy to over-annotate with these methods, in particular if using cross-species linkage, so some experimentation/review is a good idea, to tune the process for your data. Much is a judgement call.
gene Symbol from cuffdiff output
Thanks, Jen, Galaxy team