A couple questions...
I want to analyze reads of < 20bps using tophat. I see that tophat dynamically determines seed length based on the read sizes in the input file. I noticed that bowtie now supports seeds < 20bps. I have installed the latest 0.11.1 version of bowtie and the 0.10.1 version of tophat. I am able to align my short reads using bowtie and specifying a seed of < 20bps (e.g. 15 bps). However when using bowtie through tophat, tophat ends up with a seed length of 99999bp and crashes. Is there a way to specify the seed length for bowtie from tophat and avoid the crash?
Is there a way to specify the number of mismatches in the seed that bowtie uses for the genome alignment? (e.g. -n 0)
The tophat paper mentions allowing multi-reads with up to 10 hits in the genome for RPKM calculations but the default for tophat is currently 40 (-g). Was there a reason for increasing this value?
If no splice junctions are desired (just RPKMs), would one specify --no-novel-juncs and --no-gff-juncs?
I want to analyze reads of < 20bps using tophat. I see that tophat dynamically determines seed length based on the read sizes in the input file. I noticed that bowtie now supports seeds < 20bps. I have installed the latest 0.11.1 version of bowtie and the 0.10.1 version of tophat. I am able to align my short reads using bowtie and specifying a seed of < 20bps (e.g. 15 bps). However when using bowtie through tophat, tophat ends up with a seed length of 99999bp and crashes. Is there a way to specify the seed length for bowtie from tophat and avoid the crash?
Is there a way to specify the number of mismatches in the seed that bowtie uses for the genome alignment? (e.g. -n 0)
The tophat paper mentions allowing multi-reads with up to 10 hits in the genome for RPKM calculations but the default for tophat is currently 40 (-g). Was there a reason for increasing this value?
If no splice junctions are desired (just RPKMs), would one specify --no-novel-juncs and --no-gff-juncs?
Comment