Feature table import: Interactive Sample Detail versus Feature Detail

[buserj (Jessica Zandee Buser)]

Hello, and thank you in advance for your time. I am using QIIME2-2019.4 version on a conda platform.

In summary: I am looking to properly import a .tsv feature table into Qiime2, after I manually remove taxa (features) from my negative controls present in my samples. When I view my imported table in Qiime2 viewer, the tab “Interactive Sample Viewer” contains taxa identification (k__Bacteria;p__Proteobacteria;c__Gammaproteobacteria;o__Pseudomonadales;f__Pseudomonadaceae;g__Pseudomonas;s__veronii), and the tab “Feature Detail” contains my sample identification and corresponding frequencies. Due to this content switch, qiime can’t pair my metadata to my feature table anymore.

In detail: I created a relative abundance plot containing all my samples (negative controls included), and within Qiime2 view, I downloaded the .csv file of the taxa and corresponding abundances present within my samples. Within the .csv file, I manually removed the taxa present in the negative controls from the samples (if also present). For example: if taxa A occurred within my negative control 500 times, and in a sample 700 times, I would subtract 500 from 700, and 200 would be the “adjusted” abundance within the sample. After completing this for all my negative controls and samples, I deleted the negative controls, keeping only my samples and their adjusted feature abundances.

Using google sheets, I converted this .csv file to .tsv, and converted the file to .biom hdf5 file:

> biom convert 
> -i feature-table.tsv 
> -o feature-table.biom 
> --table-style="OTU table" --to-hdf5

I then imported the .biom file into qiime as a .qza file:

> qiime tools import 
> --input-path feature-table.biom 
> --type 'FeatureTable[Frequency]' 
> --source-format BIOMV210Format 
> --output-path feature-table.qza

To view this file, I converted it to .qzv:

> qiime feature-table summarize 
> --i-table feature-table-final.qza 
> --o-visualization feature-table.qzv

When visualized with the Qiime2 viewer, the file feature-table.qzv shows the taxa identification as the “Interactive Sample Detail” and my samples as “Feature Detail”. This mix-up halts any further use of this data file for diversity analyses.

Could anyone provide advice for me and my analysis?

Thank you

[thermokarst]

Hello @buserj!

What does the first few lines of feature-table.tsv look like? I suspect the table is transposed (instead of rows beings features, your rows are samples). You can transpose the table back into the “right” orientation by running feature-table transpose.

I think the more important thing here though is to point out that your negative control filtering protocol is possibly a cause for concern. This topic has been discussed extensively on this forum, but maybe @Nicholas_Bokulich or @jwdebelius would like to add something?

[buserj (Jessica Zandee Buser)]

Thank you @thermokarst,

I am really grateful for any opinions on my negative control filtering protocols.

The topics you shared have already changed the way I am thinking about working with my negative controls, and I’ll keep reading through threads within the forum. I think my next step is to assess the taxa present in the “blanks” (molecular-grade water run alongside extraction/amp/purification) and see if I can identify a pattern, rather than simply removing any taxa present in the negative controls.

Thank you!