cnr-ibba / nf-treemix Goto Github PK

• call treemix with boostrap
• create a custom R image with the packages I need (ex OptM)
• value treemix models with OptM package

more informations:

• https://github.com/carolindahms/TreeMix
• https://cran.r-project.org/web/packages/OptM/index.html
• https://cran.r-project.org/web/packages/OptM/readme/README.html
• https://academic.oup.com/biomethods/article/6/1/bpab017/6371180

sometimes treemix has issues when loading a consensus tree. It's not clear where the problem is, maybe could be enough calculate the consensus trees with dendropy and find a way to visualize this tree and its migration events. Could we define a migration consensus?

• Switch to treemix as default analysis
• Add bootstrap as option
• Divide analyses in workflows
• Calculate single migration with both softwares
• Provide already calculated trees or matrices (starts from 0 migration hypothesis)
• Document steps and pipeline execution
• Add help messages and parameters check

Hey

I contributed some years ago to a paper making a Nextflow pipeline for TreeMix. I don't give it support anymore and my git is just for the code availability of the paper, but it seems that many people are trying to use it. Unfortunately, it is based on dsl1 and mk.
So, this updated pipeline is useful for the people that have contacted me.
As the structure is already compatible with nf-core, could you add the modules and workflow properly to nf-core?
I think this way the pipeline would have more visibility and scope and people could continue supporting code.

Nice work!

• Replace treemix with OrientAGrap

Try do define a consensus tree for bootsrapped runs. I can use sumtrees from dendropy to generate a newick tree, and then call treemix by providing the calculated tree as parameter:

See here for more informations:

• https://groups.google.com/g/structure-software/c/G2Jids9wE-E
• https://github.com/carolindahms/TreeMix/blob/main/Step1_TreeMix.sh
• https://github.com/marcomilanesi/BITE