Vague Table for Taxonomic Analaysis


(Meha) #1

I used the GreenGene database to perform the taxonomic analysis. You see the first page of the result out of 25.

I am curious about the visualization file. It gave an ambiguous output. From my prospective, it is not clear which one of lines are related to what sample!

Another point is some lines consist of ONLY Bacteria kingdom. Does it mean Qiime2 is unable to detect any genus, order, family, class and phylum? Or no, there is no other levels? Is there a problem?

My third question: Will I have the same result outline with the SILVA database?

Appreciate


(Colin Brislawn) #2

Good afternoon,

I think I can answer these questions. Let’s dive in.

That’s correct. This just shows the taxonomy of different ASVs. It’s not trying to to show which samples these ASVs are in.

This means that your taxonomy assignment method was only confident down to that taxonomic level. So for that c__Clostridia near the top of the screenshot, the algorithm was confident about the kingdom, phylum, and class of this ASV (Bacteria, Firmicutes, Clostridia) but wasn’t confident about order or family so it didn’t report them.

Nope. This is normal.

Try it for yourself and find out! :wink:

Colin


(Mehrbod Estaki) #3

Hi @Mehrdad,

You are right, and that is by design. The visualization you are displaying is of the taxonomy assignments and is separate of your samples’ metadata. This file exists so that it may be used as a reference when you are performing other down-stream analysis. For example, you can use the taxa-barplot command to get sample-specific information.

This is not really a function of Qiime2’s inability to assign the taxa, but that your query sequence cannot be identified beyond that level based on the chosen database. Perhaps the sequence is too short, or the database is incomplete. Your results are very typical of this type of analysis so it is not a mistake.

Greengenes and Silva are different databases and while they should be relatively similar there are inevitably going to be some differences between the two with Silva having more sequences. In well defined communities such as human gut these database should be very similar. I know there are studies out there that compare the two so you could look for those for more details.

On a final note, I would suggest that you go through some of the Qiime2 tutorials thoroughly, and especially read Qiime2 the overview tutorial which will do a much more in depth explanation of most of the common workflows you are dealing with.
All the best.


(Mehrbod Estaki) #4

Oops, my apologie to @colinbrislawn. I didn’t mean to post over yours, probably was writing our answers at the same time again! :facepunch:


(Meha) #5

I used that but gave an error!


(Colin Brislawn) #6

For --i-taxonomy, instead of passing the classifier, pass the results you got from running the classifier. I think that’s the same file you showed in the first post.

Colin


(Meha) #7

According to your statement, I changed the i parameter content.

still have an error!


(Colin Brislawn) #8

Did you read the error?

Error: Got unexpected extra argument (taxonomylibA.qza)

I think I see where you typed taxonomylibA.qza twice. Did you notice that? :wink:

Try it again.

Colin


(Meha) #9

You are right.
But after correction I have the error!


(Colin Brislawn) #10

Thanks for trying it again!
(and thank you for posting your full error message)

Now… you tell me: what is the error you got this time? Did you find any clues about this error you can use to help us solve it?
(hint: each input to the script needs its own correct file type)

Colin


(Meha) #11

Your welcome Colin!

I used two output in two different runs:

  1. I run the command with the output I got denoised step by DADA2 called FEATURE TABLE.
  2. I run the same command but different --i-table that I put the output named taxonomylibA.qza I got from Taxonomy analysis step existing in Moving Picture platform.

As a result, I am given the same errors as you see in the two runs that’s why I sent the whole run here.

:white_check_mark: I double-checked the spelling the names. They all are ok! (Copied and pasted).

:white_check_mark: all files are required are present in the directory, so it would not be a problem as well!

:white_check_mark: In my first post in this discussion that is clear that I was successful to elicited the taxonomies but they are not in an order way to determine which one returns to which sample because of that I am going to observe them in the organised status.

In my idea, everything is ready to gain the result!
I did not troubleshoot it so far.

Thanks for replying!


(Meha) #12

Your welcome!

As you see in the photo, I run the command in two different ways:

  1. with –i-table taxonomylibA.qza which is output from taxonomy analysis that I gained from moving picture platform command.

  2. With –i-table TableDenoisedLibA.qza which is the denoised output by dada2 plugin called FEATURE TABLE that’s why I sent the whole terminal screen to be compared.

:white_check_mark: I double-checked the name of files in term of spelling.
:white_check_mark: I checked the files weather they are there or not!
:x: the problem lies on FEATURE DATA and/or FEATURE TABLE parameters!

If you have a question more, please ask me. I know them above, not further!

Thanks


(Meha) #13

Sorry I thought the first post has not been delivered then I wrote another one!


(Luca) #14

Hi Meha,
if I may give you an hint:
the errors you receiving when you run with ‘taxonomylibA.qza’ and with ‘TableDenoisedLibA.qza’ are NOT the same error.

(With the TableDenoisedLibA.qza the second case, it did NOT complain about the FEATURE TABLE parameter!)

Luca


(Meha) #15

Hi @llenzi,
Thank you for the idea!

But Franky speaking, I did not understand you completely. If you compare the horrible read scripts in both run, you will understand the both errors are returned to a requirment. In other words, the outputs require 'an argument of subtype FeatureData[Taxonomy]/ FeatureTable[Sequence]. Anyway it needs one more thing.

I followed the taxonomy analysis part of this page as I was asked:
https://docs.qiime2.org/2019.1/tutorials/moving-pictures/

qiime taxa barplot
–i-table table.qza
–i-taxonomy taxonomy.qza
–m-metadata-file sample-metadata.tsv
–o-visualization taxa-bar-plots.qzv

https://docs.qiime2.org/2019.1/plugins/available/taxa/barplot/

Every option is needed to run this command provided!

What is your opinion? Could you please tell me much more clear what is the problem?


(Meha) #16

Dear guys @llenzi @colinbrislawn @Mehrbod_Estaki

qiime taxa barplot

–i-table ARTIFACT PATH FeatureTable[Frequency]
Feature table to visualize at various
taxonomic levels. [required]
I used feature Table called TableDenoisedLibA.qza in my case!

–i-taxonomy ARTIFACT PATH FeatureData[Taxonomy]
Taxonomic annotations for features in the
provided feature table. All features in the
feature table must have a corresponding
taxonomic annotation. Taxonomic annotations
that are not present in the feature table
will be ignored. [required]

        I used the full GreenGene database (I used SILVA but gave Memorry Errorchallanging now)

–m-metadata-file MULTIPLE FILE
Metadata file or artifact viewable as
metadata. This option may be supplied
multiple times to merge metadata. The sample
metadata. [required]

    **it is so clear! I put the exact name of metadate**

–o-visualization VISUALIZATION PATH
[required if not passing --output-dir]
–output-dir DIRECTORY Output unspecified results to a directory
–verbose Display verbose output to stdout and/or
stderr during execution of this action.
[default: False]

    **They are so clear too**

I followed based on the this protocol!
I have no any misspelling either!

Now it is your turn!

thanks


(Luca) #17

Hi,
OK you getting closer!
Please go back and have a look in the tutorial where the ‘table.qza’ and ‘taxonomy.qza’ were obtained and try to understand which subtype they are.
More explicitly, is the file ‘gg-13-8-99-nb-classifier.qza’ the same subtype as the ‘taxonomy.qza’ required by the ‘qiime taxa barplot’ plug in?

Luca


(Meha) #18

I have created only ONE table so far! and the feature data is referred to reference taxonomy!
as mentioned here:

https://docs.qiime2.org/2019.1/tutorials/moving-pictures/

I confused!


(Meha) #19

Please drive me out of queue I am ganna ask another question in different area! thnx


(Matthew Ryan Dillon) #20

@Mehrdad - you are welcome to open another topic at any time. What we discourage on the forum is posting the same question multiple times in multiple topics. So, as long as your next question isn’t the same thing being asked here, you are welcome to ask now. Thanks!