Wrong of Tophat's bowtie parameters, a bug

[chenyao]

Dear all,

I use tophat --bowtie-n 4 to set bowtie mismatch, however, I found the error,

[Tue Aug 23 17:09:09 2011] Beginning TopHat run (v1.3.1)
-----------------------------------------------
[Tue Aug 23 17:09:09 2011] Preparing output location /tophat
[Tue Aug 23 17:09:09 2011] Checking for Bowtie index files
Error: Could not find Bowtie index files 4.*

It seems tophat recognize the bowtie command as a index file, however, I have pointed the index file in the command line, anyone idea?

[sdvie]

could you paste your whole tophat command here please?

[fkrueger]

Dear all,

I use tophat --bowtie-n 4 to set bowtie mismatch, however, I found the error,

[Tue Aug 23 17:09:09 2011] Beginning TopHat run (v1.3.1)
-----------------------------------------------
[Tue Aug 23 17:09:09 2011] Preparing output location /tophat
[Tue Aug 23 17:09:09 2011] Checking for Bowtie index files
Error: Could not find Bowtie index files 4.*

It seems tophat recognize the bowtie command as a index file, however, I have pointed the index file in the command line, anyone idea?

'--bowtie-n' is an option which does not take integer arguments. I guess you can specify the number of mismatches separately such as with '--initial-read-mismatches': Reads are initially mapped, allowing up to this many mismatches in each read alignment. The default is 2.

Also, -n can only be 0 to 3, and the default is 2, so -n 4 would fail as well.

[chenyao]

'--bowtie-n' is an option which does not take integer arguments. I guess you can specify the number of mismatches separately such as with '--initial-read-mismatches': Reads are initially mapped, allowing up to this many mismatches in each read alignment. The default is 2.

Also, -n can only be 0 to 3, and the default is 2, so -n 4 would fail as well.

Thank you, If it's up to 3, the I understand the error. Another question is for pair-end reads, such as 50bp*25bp, the mismatch value is the same for both reads. But since the length of reads are different, they shouldn't have the same mismatch probability. Is there anyway to seperately set the mismatch value?