How to handle samples that were sequenced twice on different runs

Sorry if this is repeated question.

I have 2x Miseq runs of identical samples that I am in the process of analysing. From what I have read the best thing to do is denoise the samples separately then merge them after confirming the results between the 2 runs are not significantly different.

Assuming they aren’t, what is the best way to merge the 2 lots of data? I have read about group and merge-tables but I’m not sure which is best for me.

And at what stage should I merge the runs? As in should I merge the feature tables then perform taxonomic classification on the combined table?

Correct!

“merge” will merge your two feature tables into a single table, and if the duplicated samples have the sample sample IDs in each run then merge will sum their read counts. If they have different sample IDs, then you can use “group” to group them together and generate a new grouped feature table.

Merge after denoising but before diversity analyses or other analyses.

Good luck!

1 Like