Hi, sometimes a few of my samples do not contain any classified reads in the kraken2 report file. As this raises an error when I want to generate feature tables with the qiime moshpit estimate-bracken command ("Error: no reads found. Please check your Kraken report"), I was wondering about a filtering functionality on kraken2 report files similar as in qiime filter samples etc. Rather than going back and filtering the demux.qza before running qiime moshpit classify-kraken2 again, it would be great to avoid long run times and computational demands of the classify-kraken2 step. What do you think? Does it make sense and is it easy to implement? Thanks!
2 Likes
Hi @Mechah - We're having some internal discussion on this and will follow up shortly. Briefly, we agree that this is important, we're just trying to figure out if anyone on the team has already made progress on it as it came up internally a while back. More soon...
Hi @Mechah ,
Thank you for the feedback and idea. We have opened an issue to get this feature added to q2-moshpit:
This should be possible and we will explore it, but I do not yet have an ETA on this feature.
3 Likes
Great! I'm looking forward to this functionality - thanks!
1 Like