Reduction in number of features After Merging and Sample Grouping!

Hi again pals,
There were two tables that I merged and grouped the samples. After merging and sample grouping, the features number went down. Please look at the following photos:

Table 1:

Table 2:

As you see, the accumulative features in both tables is 1704, while after merging and grouping the number of features dropped and reached 1254.

Merged and grouped Table:

I expected to have 1704 features but the number is reduced. Could you please justify it? what made it to be reduced?

Thanks,
Qiimer

Hi @TurboQiimer - you haven't shown us any commands that you have run - these will be very important for us to help understand what you have done to your data. Similarly, if you are able to share the visualizations, instead of screenshots, we can view the provenance, and confirm the commands run.

:qiime2:

Hi @thermokarst
Good to see you again in my topic!

Honestly, I used simple commands here you are:

qiime feature-table merge \

--i-tables groupedtable1.qza \

--i-tables groupedtable2.qza \

--p-overlap-method sum \

--o-merged-table mergedtables16sRNA.qza

For summary:
qiime feature-table summarize
--i-table mergedtables16sRNA.qza
--o-visualization mergedtables16sRNA.qzv
--m-sample-metadata-file metadata16sRNA.tsv

I have a theory in mind actually but I am not sure weather it is correct or not.

Thanks a zillion.
Qiimer

1.qzv (444.7 KB)

2.qzv (445.4 KB)

groupedtable.qzv (489.6 KB)

Thanks for sharing, @TurboQiimer - this is helpful for understanding the context of your questions. In the future, please keep this in mind when asking questions. When in doubt, provide too much information, rather than too little.

This is a good question! Why do you think that the features should be completely unique between the two runs? Are you sampling from two completely different environments? I think one way to help steer you in the right direction is to suggest you stop thinking of this as a "reduction" in features, and instead, focus on the fact that there is an overlap. This means you have some feature (ASVs) only found in one dataset, or the other, but there is also many features found in both datasets.

:qiime2:

Wow…thanks, sir. I guessed that, but I was in doubt.
Have a nice week ahead.
Qiimer

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