How can I quickly identify the ASVs belong to?

General Discussion
1

Hi, I am particularly interested in several ASVs and I have the name lists of name of the ASVs such as "9c737091c937f26a4c156b5def3291dd".

I am wondering how can I do these things, if possible can you list the name of the QIIME scripts.
I have a ASVs table of 50 samples (these samples are grouped into 5 categories)

1> How can I extract the samples with my interested list of ASVs, I would like to see the counts of each ASV in each sample.

2>How can I extract the group/category with interested list of ASVs, I would like to see the counts of each ASV in each sample.

3>How can I extract the ASVs related sequences -- I don't need the exact reads file, but I would like to know how many reads assigned to a specific ASV.

4>How can I extract the ASVs related taxonomy information. I did this in Qiime 1 where the OTU table can include the taxonomic information. If you convert the biom to excel table, the last row of this table has the taxa information. However, Qiime has an ASV table qza and taxonomy qza, when I convert the ASV to excel table, I can't find taxa information.

Thanks

3

  1. I think that to see the counts of each ASV for each sample, you want to produce a feature table. I usually get my feature tables from qiime deblur denoise-16S, but anything whose output is a qza artifact of the type FeatureTable[Frequency] (or possibly FeatureTable[RelativeFrequency]?) might be useful. Choose how to get your feature table, and I'd begin with qiime deblur --help or qiime dada2 --help.

  2. I'm not sure I understand the difference between this and question 1. But if you want to know how to merge samples that share one of your categories---is this what you want?---then you can do that with a good metadata table and qiime feature-table merge. Also, once you've solved question 1 and have a feature table, you have many options: try qiime feature-table --help.

  3. Once you have a feature table, run qiime feature-table summarize on it. View that summary qzv, with qiime tools view or through view.qiime2.org, and look at " Frequency per sample" in the "Overview" tab. It has summary statistics, and a link to download a detailed csv file if you want it.

  4. As an external resource, I've found a lot of good advice from Amplicon SOP v2 (qiime2 2020.8) · LangilleLab/microbiome_helper Wiki · GitHub. I find this problem isn't a problem for me by following their step 10, which runs biom add metadata on the biom-formatted feature table.

I hope I understood all your questions correctly, and could help at least a little bit.

3 Likes
4

Hi @sdpapet ,
Your questions have all been answered already by @timanix here:

this answers your questions 3+4 directly, and if you read the documentation at that link, you will also find that this same command works for visualizing a variety of data types: sequences, taxonomy, and feature tables, either alone or together.

Otherwise you can use qiime tools extract if you want to pass the data to other software outside of QIIME 2, e.g., for @wburgess 's suggestions here:

Good luck!

2 Likes
5

Thanks Nick. I will look at it.

6

Hello Nick,

Just to follow up.

1> I tried the hyperlink that you posted (this is website example)

qiime metadata tabulate
--m-input-file rep-seqs.qza
--m-input-file taxonomy.qza
--o-visualization tabulated-feature-metadata.qzv

Now, I can integrate my feature table and taxa table. I am wondering if I can also incorporate my mapping file into this output, where the map file has the sample and group information.

Based on your example, I only get a tabulated-feature-metadata.qzv table with all my ASV IDs and their taxonomic names, but I don't know which samples/groups they belong.

Still, if I have a ASV list, I can't filter them out or keep them in my ASV table.

2>I have list of ASV IDs that I would like to keep, which is store as a tsv file. I suppose I should use "qiime feature-table filter-features " scripts, but I can't find a good example here (Filtering data — QIIME 2 2021.4.0 documentation). I know QIIME 2 can do this, but I keep get syntax errors. Can you posted an example.
Filter a qza feature table or tabulated-feature-metadata.qzv, based on other list ASV ids.

7

sounds like you want to work with an annotated biom-format table, not a QIIME 2 feature table (which stores feature count data separately from feature and sample metadata). You should work with biom-format if you want to merge these data and metadata together on multiple axes.

you can add a feature table as an input file to metadata tabulate to see the sample/count information alongside the feature data (note: you will need to transpose the table first with qiime feature-table transpose). But this will only contain sample IDs, not group IDs (unless if you use qiime feature-table collapse first).

yes. You would use that command, and input your list of feature IDs as a metadata file to that command.

For more support with that or other questions, open a new topic thread — asking multiple questions in a single thread makes it difficult for moderators to moderate and readers to read. Thanks!

1 Like
8

Hi Nick,

Thank you very much.

1> You are correct. The best way to work on this is using the biom-format ASV table with taxonomy information. However, I don't know how to do this, because currently the ASV information and taxonomy information are stored in two qza files.
For example, if I run this qiime tools export --input-path demux.table.qza --output-path .

I only got biom file for ASV counts, after I convert biom to txt file?

Do you know I can I get a txt/csv file converted from biom file which from ASV feature table and taxonomy table

Can you give me an example, which scripts I should use?

I need a excel spread sheet with both ASV counts and taxonomy information, but I don't know I can I do this in QIIME2.

2> I have an additional questions about the sample metadata file (aka, mapping file).

As you know, we use this initially, when we build an ASV table.

Here is I am going to do. When I do the down stream analysis, for example, PCoA analysis, or summarize the taxonomy. I would like to change a group scheme. So I need to revise the original mapping file (e.g., I moved 10 samples from group A to group B).

Do I need to rebuild the ASV feature table from the very beginning? For example, rebuild the ASV table, subsample the table, filtering the unwanted ASVs etc.?

Or I can use the old ASV feature table. I only need to use my new sample metadata file to build PCoA plot or summarize the taxonomy to build a bar plot.

Thanks,

9

I described how to do this above. You can combine metadata on a single axis (feature data OR sample data) but not both. To do that you need to export each artifact individually and use biom-format. QIIME 2 does NOT store feature count data together with taxonomy and sample metadata in a single artifact.

2 Likes