Linking Greengenes2 to another feature table

User Support
,
1

I have generated the FeatureTable[Taxonomy] and FeatureData[Taxonomy] using the v4 code. I used qiime metadata tabulate for both to generate two tables. I have two questions:
1- Does the table generated from FeatureTable[Taxonomy] show the count data for each feature vs ID? I want to export to R to do finish the stats there so need the count data
2- The table from FeatureData[Taxonomy] shows the confidence intervals for all features as 1. I had this issue when using the non-v4 code and it was answered in a previous question.
Thanks, any help would be appreciated.

2

Hi @newberrf,

For (1), you need to construct a FeatureTable[Frequency] to have count information.

Is there a question for (2), it sounds like you tracked down a prior answer?

Best,
Daniel

3

Hi @wasade, thanks for the quick response

  1. Okay, I'm an idiot. greengenes2 filter-features generated a FeatureTable[Frequency], not a FeatureTable[Taxonomy] like I said in the post. The qzv looks like this:
    image
    image1553×289 15.5 KB

Which is count data, right? I am doing some complex statistics in R and it is better to have the count data than continue with the stats offered in qiime2.

  1. I thought the reason the confidence intervals for the non-v4-16s code was 1 was due to it being a wrapper for q2-vsearch and the taxonomy being fixed. (so value set to 1). I'm just concerned that I have used the v4 code and still getting a confidence interval of 1 for all features. I'm doubting the reliability of my work. I'm new to qiime2 so this is a steep learning curve.

Thanks

4

Hi @newberrf,

None of this stuff is easy :slight_smile:

That looks like a feature table although I'm used to seeing the features on the first column rather than the first row. The confidence values though would be associated with FeatureTable[Taxonomy] not FeatureTable[Frequency].

filter-features is using existing placements, which also implies that the ASVs being matched have been observed previously.

Best,
Daniel

5

Thank you, appreciate the response.

Sorry for you mind explaining what you mean by the last sentence? I'm used to doing 16S analysis in vegan and having to triple check everything I am doing in qiime2.

6

filter-features performs a set intersection with features that exist in Greengenes2. The previously placed ASVs are derived from samples indexed in Qiita. Does that make sense?

Best,
Daniel