Advice on alternatives to CIPRES for population structure analyses
Anaïs Beaumariage
I am currently using the Parallel Structure tool on CIPRES to perform population structure analyses with STRUCTURE. I know that CIPRES will cease operations on March 31, 2026, and I wanted to ask for advice regarding possible alternatives.
I am able to run STRUCTURE locally through RStudio, but this setup is not well suited for large datasets or computationally intensive analyses, which is why CIPRES has been essential for my work.
In particular, I was wondering:
- Are there other platforms or services that offer similar capabilities for running STRUCTURE or Parallel Structure?
- Is it possible to run the same program on another computing cluster (e.g. institutional computer), and if so, are there recommended setups or resources?
- Do you have any documentation or best practices for migrating analyses currently performed on CIPRES to another system?
Any guidance or recommendations would be greatly appreciated, as this tool is central to my current analyses.
Thank you very much for your time and help.
Anaïs Beaumariage
Miller, Mark
Hi Anais,
Thanks for your message. Let me say first that we can announce that the resource will be open an extra three months, till 6/30/2026.
I am unaware of any other resource offering these codes via public access.
We would provide the CIPRES infrastructure code to anyone who asks, but the overhead of setup will be significant even for a highly skilled developer.
It is certainly possible to run this on any cluster, from the command line.
The instructions for doing so are part of the code distribution.
The main thing is having some knowledge of submitting run scripts to a batch queue, and what resources you have available.
We have optimized the code for our local resource to insure most parallel bang per processor on the jobs submitted to CIPRES.
One thing I can say about ParallelStructure is it is research code meaning the error messaging isn’t fully developed, so trouble shooting can be a bit tricky.
However, if you are running on CIPRES already, you shouldn’t have a problem.
The optimization for parallel processing is cluster specific, and will have to be repeated on your resource, if your data sets have a lot of variability in size and complexity.
We may be able to provide you with some bench marking test cases should you require them.
I will leave it to my colleague Wayne Pfeiffer (our HPC expert) to comment on any specific tips he might have for this code.
Sincerely,
Mark MIller
--
You received this message because you are subscribed to the Google Groups "CIPRES Science Gateway Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
cipres-science-gatew...@googlegroups.com.
To view this discussion visit
https://groups.google.com/d/msgid/cipres-science-gateway-users/dcfc17d8-e796-4004-a422-81dacff74e92n%40googlegroups.com.
