Can I merge taxonomy results made with different q2 versions or coming from different gene regions?

Hi,

I have data that is coming from multiple sequencing runs. The sequencing was done either for the V4 region or for V3/V4. Usually I merge new datasets into older datasets that are already merged. I update Qiime2 with each new release and always use the latest version. This means that the data is based on multiple Qiime2/Dada2 versions. After merging the new data into the old data, I run the taxonomy again.
Overall I keep the data from the two different regions separate in Qiime2.
For the taxonomy I use qiime feature-classifier classify-sklearn with the provided Silva classifier.

Now I have the following questions:

  • Can I first run the taxonomy per dataset/sequencingRun and merge the files afterwards even if the single taxonomy results are based on different Qiime2/classifier versions? I think I read somewhere that the different versions are not compatible but I can’t find it anymore…
  • Can I merge the taxonomy for the V4 data with the taxonomy for the V3/V4 data or is there anything I need to consider?

Thanks in advance!

Best regards,
Thalla

Hi @Scymna,

Typically, the best approach is to run a consistent pipeline, even if it isn’t the newest. I recommend reprocessing all your data again usisng consisteent parameters. Theoretically it really shouldn’t matter. Theoretically is a pain when you’re writing a methods section.

What i tend to do (depending on my avalaible computational resources) is:

  • Run each sequencing run through DADA2 using primer pair specific consistent parameters (if Im mixing primer pairs)
  • Merge my feature tables at the ASV level and my esquences
  • Classsify taxonomy on all my sequences (more computationally effecient, only one taxonomy artifact to handle)
  • Build a phylogenetic tree with all my sequences (no need to rebuild)

As far as the second issue goes - my recommendation here is to consider close reference OTU clustering for mixing regions. Your ASVs won’t overlap in this meta analaysis, they will start at different places and be different lengths. There are tricks you might be able to play to try and merge them - but they’re not tricks that are currently easily in QIIME 2.

Best,
Justine

4 Likes

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.