Hi Ben, Back in version 2.0.1, I added a feature that I hope addresses this issue: -- Added --non-deterministic option, which better matches how some users expect the pseudo-random generator inside Bowtie 2 to work. Normally, if you give the same read (same name, sequence and qualities) and --seed, you get the same answer. --non-deterministic breaks that guarantee. This can be more...
2013-01-29 14:09:51 PST in Bowtie
Thanks very much for this suggestion. I fixed this in v2.0.5. Best, Ben.
2013-01-29 14:08:29 PST in Bowtie
This is an excellent point. I added a note to the manual in v2.0.5 indicating that this is the case. It's an inherent drawback of the method used to search for the opposite mate in Bowtie 2. Thanks for this suggestion, Ben.
2013-01-29 14:07:48 PST in Bowtie
Re: "Is there any reason to think that the resulting alignment files produced by 2.0.6 would be different from those produced by 2.0.4 in some systematic fashion, using the exact same command?" No, I can't think of any reason. I'll close this issue for now. Best, Ben.
2013-01-29 14:02:01 PST in Bowtie
Hi, This is a discordant pair (extra flag YT:Z:DP). So this is expected. Here's how the manual describes how Bowtie 2 fills in the TLEN field: -- Inferred fragment length. Size is negative if the mate's alignment occurs upstream of this alignment. Size is 0 if the mates did not align concordantly. However, size is non-0 if the mates aligned discordantly to the same chromosome. --...
2013-01-29 14:00:20 PST in Bowtie
Thanks for this very complete report! If you can, could you try again with Bowtie 2.0.6 and let me know if this is still occurring? Best, Ben.
2013-01-27 17:45:01 PST in Bowtie
ben_langmead made 3 file-release changes.
2013-01-27 12:47:03 PST in Bowtie
ben_langmead made 2 file-release changes.
2013-01-27 12:44:03 PST in Bowtie