Conda repository changes affecting local installs of earlier versions of GATK (Forum post included)

[gokalpcelik]

Hi
We have a forum post asking help for getting GATK 4.1.0.0 conda environment installed using the yml file.
https://gatk.broadinstitute.org/hc/en-us/community/posts/18332470602523-Install-GATK-version-4-1-0-0-using-Conda-

Looks like restructuring of the default repository under conda took out some of these packages and they are no longer directly accessible. They can be accessed from the forge repo with certain flags. This issue seems to deprecate some of the older but still usable versions of GATK (due to various reasons). Directing people to use the docker version or upgrading to the latest GATK version seems to be the only solution left for now.

Any other ideas of how we should pursue this issue? @lbergelson @droazen ?

[samuelklee]

Thanks, @gokalpcelik! That's unfortunate that the packages are no longer accessible. However, this old version of the environment was also somewhat problematic in that it included build strings, installed many packages using pip instead of conda, and was otherwise overly restrictive, see e.g. https://gatk.broadinstitute.org/hc/en-us/community/posts/360061666671-Broken-conda-env-create-n-gatk-f-gatkcondaenv-yml

I would suggest, in the following order:

1. Users upgrade to more recent versions of GATK
2. Users use the 4.1.0.0 Docker
3. Users edit the 4.1.0.0 environment to be less restrictive. For example, using Ubuntu 20.04.2 + conda 23.10.0, I was able to build something that is probably sufficiently close to the 4.1.0.0 conda environment by making the following changes:

< - openssl=1.0.2l=0
< - pip=9.0.1=py36_1
< - python=3.6.2=0
< - readline=6.2=2
< - setuptools=36.4.0=py36_1
< - sqlite=3.13.0=0
< - tk=8.5.18=0
< - wheel=0.29.0=py36_0
< - xz=5.2.3=0
< - zlib=1.2.11=0
---
> - openssl=1.0.2l
> - pip=9.0.1
> - python=3.6.2
> - setuptools=36.5.0
> - wheel=0.29.0
> - xz=5.2.3
> - zlib=1.2.11 

There is some clobbering that happens with the pip installs, but nothing major. Note that I cannot guarantee that this environment exactly reproduces results generated with the Docker (and it's possible that there might be a more minimal fix), but it's perhaps a starting point for those users that cannot go with the other (highly preferred) options.