My guess is that other plugins or packages were installed in that environment, leading to the conflict. RESCRIPt should not have issues installing in a fresh 2023.2 environment (unless if some dependency packages were changed on the conda side, which could happen).
You should try installing in a fresh environment, or create a minimal RESCRIPt environment (following the instructions on GitHub) to avoid dependency conflicts with other plugins that you have installed.
Let me know if you still experience issues with a fresh environment or minimal environment! Good luck!
Thanks for the suggestion. I tried the minimal install and I got a different error message. It appears to not be able to find a lot of neccessary packages. From the channels it looks in but finds nothing, could this be an issue with RESCRIPt and compatibility with the M1 Conda install?
"PackagesNotFoundError: The following packages are not available from current channels:
Are you sure your in the "minimal install" rescript environment. I often see the errors you've posted when I am accidentally trying to run the conda install ... commands in my base or other environment.
I just ran each of these separately on my M1 MacBook Pro, and everything worked as advertised
Following your suggestion, I removed the old RESCRIPt minimal environment I had installed and tried reinstalling following the instructions from GitHub.
Unfortunately, I get the same error message as I posted previously
Hi @aghudson ,
It looks like you are having issues more broadly with conda and with installing QIIME 2 in general, not just RESCRIPt:
So let's close this topic and wait for the other to resolve. If you are still having issues with RESCRIPt after that other topic is solved, please feel free to re-open this. But for now I would consider these duplicate posts, as they almost certainly stem from the same issue.
I think I have come up with a solution with the overall Qiime2 installation (please see my reply in the thread). I havenʻt tested it extensively, but the installation ran to completion and I can load the Qiime2 conda environment without any error messages. I obviously canʻt rule out that this is not still causing the ongoing issues with RESCRIPt though.
Hi @aghudson ,
Just to clarify: are you getting the same error message as before, or has the message changed now that you have QIIME 2 working?
One more idea: you could try using mamba instead of conda, it tends to solve environments much faster than conda (or at least might give a clearer error message)
For the minimal RESCRIPt install, I get the same error message posted above.
For the Qiime2 environment installation, I am not sure. I only see a portion of the error message in the terminal, which is just a long list of the packages that are showing conflicts. Many of the conflicts I can see in the latest attempt, seem to be related to R or R-bioconductor.
We are not able to reproduce this error, unfortunately, including @SoilRotifer who is also using an M1 mac. So I think that your conda configs might still be messed up.
I basically gave up with my M1 mac and tried installing everything on an older Intel Macbook that has never had Qiime2 installed on it, though it has a version of Miniconda installed.
I followed the Qiime2 installation instructions for Intel Mac (version: qiime2-2023.2), including updating conda. Everything ran/installed ok. In the Qiime2 conda environment I had just created, I attempted to install RESCRIPt.
I ran the following command to install dependencies:
Unfortunately, I get pretty much the same error message. The weird thing is that a lot of the listed package conflicts are the same for the old and new macs, which I wouldn't have thought would be the case if this was to do with my current conda configuration. I can send the portions of the error messages I was able to save for both computer installations, I am just not sure what the best method of doing that might be.
I get the feeling that most of the conflicts are to do with the q2-types-genomics package as I can install the other 2 dependencies separately, but I can't prove that.