Error in using uploaded Experimental Conditions

[Adrian-Zet]

Dear GEOexplorer team,

I get the following error when I try to upload an experimental conditions file (for both Microarray & RNASeq Data). I followed the exact structure present in the template for both yet it always pops this error no matter what I try:

"There was an error incorporating the (first) experimental conditions file. Please ensure it is in the same format as the example file in the 'Example Datasets' tab"

Update: I managed to solve the issue for the RNA-Seq Data. It was an error on my part for the input. Yet I didn't manage to solve it for Microarray data.

Here are the experimental expression file:
gse-expression-profile.csv

And here is the experimental conditions file:
GSE73002_sampleInfo-subsetted.csv

[guypwhunt]

Good afternoon

Thank you for bringing this to our attention.

I think if you use this expression dataset it should work (I just changes the gene expression values from general to numerical in Excel).
gse-expression-profile.csv

If using GEOexplorer is too slow in processing the data let me know and I can run the analysis using the back end functions.

I can see you have thousands of samples and it might be very slow running it on GEOexplorer.

[Adrian-Zet]

Thank you very much!

Indeed, changing the type from general to numerical solved it. I will do this step from now on.

I have just another question:
In the Exploratory Data Analysis the contrast colouring is made as a different color for each GSM. Is there any way in which I can perform the PCA or UMAP using the same groups that I set during Differential Expression Analysis? Or do I need to use the backend code and modify it for that?

Edit: I attached a picture with an example, I would like to know if instead of having one color for each GSM I could have one color for Group 1 and another for Group 2:

Thank you again for helping me with the previous issue and the fast answer!!

[guypwhunt]

That's no problem at all.

I have actually been working on that update recently. But it will still take a little time to incorporate it into the website.

So I can either create those plots for your or send you the code that allows you to plot them yourself. Whatever suits you best.

I assume you want something like this:

[Adrian-Zet]

Thank you again for the very fast feedback!

If you could send me the code to make the plots myself until the update is done it would help me a lot. I prefer it since I will need to analyze some other sets too.

The way you assumed is the exact way I need it.

Have a great day!

[guypwhunt]

That is no problem at all.

Below are scripts to run the analysis and produce the plots etc.

I have set the scripts to run from the microarray dataset you sent.
adrianRScripts.zip

Let me know if any of it doesn't make sense or if you have any questions etc.

[Adrian-Zet]

Thank you very much again for all the help and support!

I have one last question related to the package. Do you plan to add support for PCR data in the future or Nanostring? Or would it work if I tried to use a PCR dataset on the Microarray setting?

Once again thank you for the amazing feedback, please close the issue if you see fit.

Have a wonderful day!

[guypwhunt]

That is no problem at all.

Do you have an example PCR dataset? Apologies I have never worked with PCR data before.

I suspect GEOexplorer will work fine with the PCR dataset. GEOexplorer is built upon packages that are pretty universal for this kind of analysis. E.g. I use GEOexplorer for some lipidomics and proteomics projects I am doing.

The only things I want to check are if the PCR dataset should be arcsine transformation instead of log-transformed and if it should be treated as microarray or RNA-seq. But I should be able to tell pretty quickly if you have an example dataset that I could see.

[Adrian-Zet]

GSE41922.csv
GSE41922_sampleInfo_subset-check.csv

I used this dataset to check GEOExplorer on PCR data.

They are already normalized and log2 transformed as I checked the paper. I guess in this case arcsine transformation is no longer necessary (is arcsine more widely used on PCR data than log2?)
I tried using the Microarray setting and it seemed to work flawlessly.

I attached them here in case you want to test them yourself.
I will also try the RNA-Seq setting and see if anything changes.

Thank you for your involvement and interest!

[guypwhunt]

That's no problem at all and thank you for sharing the dataset.

Using the microarray setting will be perfect for the PCR datasets.

I did some digging and it is far more common to log transform PCR data than arcsine transformation. So forget I mentioned arcsine transformation.

Feel free to reach out if you need any more help/advice.

Have a lovely weekend.

[Adrian-Zet]

Dear Guy,

I hope you are doing well.
I come back with a question related to the local installation of the GEOExplorer package.
I haven't used it in a while and I noticed that my local installation doesn't have the option to upload a DataSet (both expression & metadata as the online server allows) made by me (only allowing using data available from GEO).

Is this the intended way the package works or am I using an older version? I tried reinstalling it but it didn't offer the option.

Or for the local installation is just the back-end script the one that allows using your own datasets?

Best regards,
Adrian.

[guypwhunt]

Hi Adrian

Nice to chat with you again.

Hmmm, I will have to look into it.

In the meantime, you can use this code to install a local version from my other GitHub account.

library(devtools)
install_github("KHP-Informatics/GEOexplorer")

library(GEOexplorer)
loadApp()

I have checked and it should enable you to manually upload gene expression files and metadata.

All the best,
Guy

[Adrian-Zet]

Thank you for your prompt answer Guy!

I tried to install the software with the commands you provided and I got the following errors:

`Error in loadNamespace(j <- i[[1L]], c(lib.loc, .libPaths()), versionCheck = vI[[j]]) :
namespace ‘tidyr’ 1.2.1 is already loaded, but >= 1.3.0 is required
Calls: ... namespaceImportFrom -> asNamespace -> loadNamespace
Execution halted
ERROR: lazy loading failed for package ‘ggpubr’

• removing ‘/export/home/math/saia/R-4.2.1/lib/R/library/ggpubr’
• restoring previous ‘/export/home/math/saia/R-4.2.1/lib/R/library/ggpubr’

The downloaded source packages are in
‘/tmp/user/1233/Rtmp1BTmdB/downloaded_packages’
Updating HTML index of packages in '.Library'
Making 'packages.html' ... done
── R CMD build ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
✔ checking for file ‘/tmp/user/1233/Rtmp1BTmdB/remotes14a92e6bbd2751/KHP-Informatics-GEOexplorer-d34ff61/DESCRIPTION’ ...
─ preparing ‘GEOexplorer’:
✔ checking DESCRIPTION meta-information ...
─ checking for LF line-endings in source and make files and shell scripts
─ checking for empty or unneeded directories
─ looking to see if a ‘data/datalist’ file should be added
Omitted ‘LazyData’ from DESCRIPTION
─ building ‘GEOexplorer_1.2.0.tar.gz’

• installing source package ‘GEOexplorer’ ...
  ** using staged installation
  ** R
  ** inst
  ** byte-compile and prepare package for lazy loading
  Warning: replacing previous import ‘shiny::dataTableOutput’ by ‘DT::dataTableOutput’ when loading ‘GEOexplorer’
  Warning: replacing previous import ‘shiny::renderDataTable’ by ‘DT::renderDataTable’ when loading ‘GEOexplorer’
  Error in loadNamespace(i, c(lib.loc, .libPaths()), versionCheck = vI[[i]]) :
  there is no package called ‘genefilter’
  Calls: ... loadNamespace -> withRestarts -> withOneRestart -> doWithOneRestart
  Execution halted
  ERROR: lazy loading failed for package ‘GEOexplorer’
• removing ‘/export/home/math/saia/R-4.2.1/lib/R/library/GEOexplorer’
• restoring previous ‘/export/home/math/saia/R-4.2.1/lib/R/library/GEOexplorer’
  There were 17 warnings (use warnings() to see them)`

Can you please help with this issue?
I'm not sure if it is related to the R installation or if it has to do with the dev version of this GEOExplorer package.

Thank you for your help!
Adrian.

[guypwhunt]

Hi Adrian

That is no problem at all and I am sorry the installation didn't work.

I checked that GEOexplorer installs okay on a fresh version of R, with no other packages pre-installed. So it might be something to do with the ggpubr and tidyr.

Hopefully, manually installing/updating ggpubr and tidyr should fix the issue.

If you run this code to install the packages:

install.packages("ggpubr")
install.packages("tidyr")

Then try to install GEOexplorere again:

library(devtools)
install_github("KHP-Informatics/GEOexplorer")

library(GEOexplorer)
loadApp()

Let me know if this doesn't work and maybe we can share screens to try to debug what is going on.

All the best,
Guy