please answer a rapid question!

Is the sequence count that is generated from q2_feature_table.summarize, is the number of OTU for each sample???

Not exactly!!! It is the number of reads in that sample!!!


As a follow-up here, I’m a little bit sorry for the tone in the last response, I intended it as a playful response, but it could certainly be read differently. The internet never transmits tone the way you intend.

That said, it can be sometimes be frustrating to feel like our time is taken for granted, even though it is a considerable investment on our part. So we appreciate it if posters can take the time to clarify what specifically they are looking at as it took me a few minutes to work out what you were talking about (and I’m not even 100% sure I found the same thing).

As a more relevant aside, we’ve recently changed the language from Sequence Count to Feature Count which is hopefully less confusing.

Let me know if you have any other questions about this! I’m happy to help :slight_smile:


So, can i know the OTUs per sample number? Also, i have another query what is the difference between number of OTUs and observed OTU at specific sampling depth?!

Also, is the feature count indicating for anything like, diversity, richness or number of bacteria?

Hi @Soha,
Just a reminder that as per the forum’s cod of conduct you should ask new/non-related questions in a separate thread. This will help with keeping the forum organized and helps others searching for relevant questions easier.
I would also recommend going over one of the many tutorials available in Qiime2, for example starting with the Moving Pictures tutorial as it covers most of your inquiries.

In short, the number of OTUs/features per sample is the sample’s ‘richness’ which can be calculated using the observed_otus metric in the qiime diversity alpha plugin.

Sampling depth is the value where all your samples are sumsampled/rarefied to without replacement before calculating diversity metrics. Without it your diversity metrics are calculated based on all the reads of your each sample, ignoring the issue of uneven sampling depth.