Looking for recommendations for downstream analysi...
# taxprofilerm
Megan Justice
10/21/2025, 5:42 PMLooking for recommendations for downstream analysis tools for statistical comparisons and grouping of samples after the Bracken step. Any suggestions?
j
James Fellows Yates
10/22/2025, 8:40 AMThis depends a lot on what the end goal is... normally my recommendation here is to find papers doing similar things you want to do and replicate their methodlogy
m
Megan Justice
10/22/2025, 5:39 PMThanks, James! I'll get some clarity on what I need and do some literature searching! 🙂
Megan Justice
10/22/2025, 6:41 PMI'm also curious if you know how to best introduce metadata into the pipeline? I have the output for each sample individually and would like to pool/average samples that are from the same condition. That would allow me to make more broad comparisons in things like Krona instead of viewing one sample at a time. I'm also hoping to get statistics on my groups, like diversity metrics. I found the KrakenTools to allow me to conduct the analysis, but I think it is only on samples not groups.
m
moritz
10/23/2025, 7:10 AM@James Fellows Yates this is the kind of thing I have in mind under data management.
@Megan Justice out of curiosity, how did you solve your AWS challenges?
For this problem, I would combine all information in a TreeSummarizedExperiment in R with the mia package. They are about to publish a guide as well.
👍 2
🙌 1
m
Megan Justice
10/23/2025, 3:43 PMThanks so much for the suggestion!
As for AWS, I wasn't ever able to get the Taxprofiler running all the way through with it. It ended up being smoother and faster if I utilized my university's HPC system, specifically the big memory nodes they have available. To keep troubleshooting on AWS was getting costly so I have switched gears to the 'free' system on campus.
👍 2