jtamames / SqueezeMeta

A complete pipeline for metagenomic analysis
GNU General Public License v3.0
372 stars 80 forks source link

SQMtools version #729

Closed HannahBenisty closed 1 year ago

HannahBenisty commented 1 year ago

Hi,

I am trying to run SQMtools but I have a problem with the versions and I am not able to load the tables. "Your project was created with SqueezeMeta v1.5, while this is SQMtools v1.6.2. You can ignore this message if things are working fine for you, but if you experience any issue consider using the right version of SQMtools for this project". So, I would like to use the correct version of SQMtools. How can I do it?

Thanks.

fpusan commented 1 year ago

If on windows/Mac, follow the install instructions here https://github.com/jtamames/SqueezeMeta/wiki/Using-SQMtools-(pre-1.6.2)-in-a-Windows-environment You need to make sure to download the tarball for SqueezeMeta v1.5

If on linux, just install SqueezeMeta v1.5. via conda, the right SQMtools version will be shipped within the environment

HannahBenisty commented 1 year ago

Thanks! I am on Mac and I have followed all the steps. Could please help with this one: 1-Download the file SQMtools - For instance, go to downgit and put the URL https://github.com/jtamames/SqueezeMeta/tree/master/lib/SQMtools to download the package as a zip file. Which would be the link I should use to download the SQMtools from v1.5?

In parallel I have been trying to install the package directly in RStudio, but I keep getting this error:

install.packages("~/Downloads/SqueezeMeta-1.5.0-post2.tar.gz", repos = NULL, type = "source") Warning in untar2(tarfile, files, list, exdir, restore_times) : skipping pax global extended headers ERROR: cannot extract package from ‘/Users/hbenisty/Downloads/SqueezeMeta-1.5.0-post2.tar.gz’ Warning in install.packages : installation of package ‘/Users/hbenisty/Downloads/SqueezeMeta-1.5.0-post2.tar.gz’ had non-zero exit status

Thanks again

HannahBenisty commented 1 year ago

Hi,

I have managed to install the library from the v1.5.0 version. However, when loading the tables in RStudio I get this message:

Error in SQM$functions[[method]]$tpm["Unmapped", ] : subscript out of bounds In addition: Warning message: In sprintf("%s/results/tables/%s.superkingdom.%s.abund.tsv", project_path, : one argument not used by format '%s/results/tables/%s.superkingdom.%s.abund.tsv'

I have used the same structure and names in my local computer. Do you know where could be the problem?

Thanks

fpusan commented 1 year ago

Did you run sqm2tables.py using the same environment in which you ran the main pipeline?

fpusan commented 1 year ago

Just in case, re-run sqm2tables.py making sure that you have SqueezeMeta 1.5.0. Test that the project can be loaded in your server using the SQMtools shipped with that version. If everything works, then copy the project to your laptop and try analyzing there. Alternatively try repeating the analysis with a newer version, we have simplified how SQMtools is installed in WIndows/Mac, and the transfer of projects between different computers.

HannahBenisty commented 1 year ago

Thanks, I will try to re-run.

Alternatively, as you suggest, I will repeat with the newer version. You mean Squeezemeta 1.6.3?

Thanks

fpusan commented 1 year ago

Well 1.6.3 is not officially out yet, if everything goes fine I will release it by the end of this when week/beggining of the next.

HannahBenisty commented 1 year ago

Great, thanks! Then, I will wait for it.

HannahBenisty commented 1 year ago

Hi, I have re-run my sample with the new version. Could you please indicate me what would the steps I should follow to use SQMtools on these results (with Mac). Should I still follow this steps that you recommended? https://github.com/jtamames/SqueezeMeta/wiki/Using-SQMtools-(pre-1.6.2)-in-a-Windows-environment

Thanks!

fpusan commented 1 year ago

No, it's easier now. See https://github.com/jtamames/SqueezeMeta#15-analyzing-squeezemeta-results-in-your-desktop-computer

HannahBenisty commented 1 year ago

Thanks, everything worked perfect.

fpusan commented 1 year ago

Glad to hear!