Hi Mingquan,
Please send your queries to one of our mailing lists -- galaxy-dev or galaxy-user -- rather than to individuals. That way, more people can see and help you with your questions, you'll likely get a more timely response, and the mailing list archive can serve as a useful repository. Your question about GFF to BED is best addressed to galaxy-user (b/c it's about usage of Galaxy), and your question about Tophat is best addressed to galaxy-dev (b/c it's about setting up a local Galaxy instance). Hence, I've cc'd both lists.
Ok, on to your questions:
> Maybe i can use the tophat anlone.
We've got a Galaxy wrapper ready for Tophat version 1.1+ ; it will likely be available in galaxy-central this week. Once you update your version of Galaxy, you'll be able to run Tophat in Galaxy.
> But now i have a another question for giff to bed function.
> i want to use gff-to-bed function after i run it, it seems that i got the results :
>
> empty, format: bed, database: ?
> Info: 0 lines converted to BED. Skipped 74166 blank/comment/invalid lines starting with line #1.
>
> i am using gff3 file.and it is the first few lines.
>
> ##gff-version 3
> ##genome-build MSU Rice Genome Annotation Project osa1r6
> ##species Oryza sativa spp japonica cv Nipponbare
> ##sequence-region Chr1 1 43268879
>
> Chr1 MSU_osa1r6 gene 1903 9817 . + . ID=13101.t00001;Name=TBC%20domain%20containing%20protein%2C%20expressed;Alias=LOC_Os01g01010
> Chr1 MSU_osa1r6 mRNA 1903 9817 . + . ID=13101.m00001;Parent=13101.t00001;Alias=LOC_Os01g01010.1
>
> Chr1 MSU_osa1r6 five_prime_UTR 1903 2268 . + . Parent=13101.m00001
> Chr1 MSU_osa1r6 five_prime_UTR 2354 2448 . + . Parent=13101.m00001
This appears to be a valid GFF3 file and should work fine. We'll look into this and get back to you.
Thanks,
J.