Issue With Groups Tab

[grajimen]

Hello,

I am not sure how to use the group selection on the right-hand side of the Groups tab. I have 9 organisms within 10 samples, and I am looking for differential expression between organisms within the samples. I have tried adding the sample name and the organism name on the left-hand side (Group Names) but I do not know how to determine which value to assign to each group. What is this value that we are required to input?

Thank you for your help!!

[jcooperdevlin]

Hi,
How many samples do you have total?
The differential abundance calculations require at least two samples per group in order to have sufficient power to perform the analysis. Additionally, we limit the number of groups to ten total to prevent overloading of the server.

[grajimen]

Hi thank you for getting back to me! I have 10 samples total with a mixture of organisms in each sample (there are 9 possible organisms we considered). I did see errors for the differential expression analysis but wasn't sure if it was due to the input file or the study design/sample size. I will discuss the setup with my PI to see if we have additional samples to include for DE analysis.

Another thing I noticed after looking over requirements again for Wham, I can see that my input file format is not the recommended format of Acc Feature Taxa (as first three columns of TSV file), but I used the recommended python file conversion script. After reading over the help file for humann_2_wham.py, I think my input matched the format of the humann2 input and output matched the format of output created by humann_2_wham.py of Acc GeneFamily SampleAbundanceValues (as first three columns of TSV file). Since this is still missing Feature and Taxa, I was not sure if I should be using another script as well before at least creating the abundance plot.

[jcooperdevlin]

Happy to help with this!

It sounds like the 10 samples you have do not fall under any sort of groups (controls versus experimental condition) and that each one is an independent sample. Therefore the differential calculation will not work because there is only one sample per group. Despite there being nine possible organisms per sample the current design of your samples does not allow for differential abundance determination, unless you can group your samples based on some experimental criteria (treatment, source etc.).

As for the formatting could you send the first few lines of your input file and I will have a look. Depending on the humann2 output the conversion scripts can behave a bit funny.

[grajimen]

Thank you so much for your help. That certainly makes sense that the differential expression calculations are not intended for a design like this.

I have attached the tsv file with the first few lines of data. It looks like I can I input the data into wham, but my Groups page appears differently than it appeared in the article discussing wham. I enter group names on the left and then the right-hand side has an equal number of boxes where I am required to select a value from the dropdown (e.g. 0,1.5x10-6,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8 are provided in the dropdown on the right if I choose to have 10 groups).

Thank you!!

humann2_genefamilies subset.txt

[grajimen]

update: I attached the wrong table in my previous comment. That was the format I had before running the python conversion script. I have attached the tsv file with output from the wham conversion script that should be input for wham app.

wham_genefamilies_named subset.txt

[jcooperdevlin]

Sounds like strange behavior! Looking at your file it appears there is an extra new line character between the first line and the second line so try deleting that.
Additionally, it may also be due to the sample names starting with a number, try changing them to start with a letter.

[grajimen]

Thank you for that suggestion, it seems to have worked! After deleting that extra line and renaming sample columns to remove numeric prefixes, the format is working for the app. I am now able to specify samples that belong to each group on the right-hand side of Groups tab instead of choosing values from a dropdown menu on the right. Thanks!

[grajimen]

Hello, I am running into the same issue as before where I upload my humann2 genefamilies tsv file into the app and the groups tab gives me a dropdown menu of numeric values instead of my sample names. I try to fix the issue as I had previously (by starting sample names without any numeric prefix and removing any additional blank lines between header row and subsequent data rows). However, I keep receiving a message that the file uploaded immediately followed by an error message 'Disconnected from server'. I did not change any of the steps I took since the last set of samples I was able to run, and I am wondering if you have any suggestions on addressing this issue.

Also, I do not receive the same error 'Disconnected from server' when I run the genefamilies tsv file for my last group of samples. I have attached a subset of the non-working file (genefamilies_analysis_two) as well as a subset of the working file (genefamilies_named).

wham_genefamilies_analysis_two.subset.txt

wham_genefamilies_named.subset.txt

Thank you in advance for any help/input!