search

cole-trapnell-lab / cicero-release

https://cole-trapnell-lab.github.io/cicero-release/
MIT License
56 stars 14 forks source link

Transfering mapping info from one object to another #69

Closed aleksandrPi closed 3 years ago

aleksandrPi commented 3 years ago

Hi monocle/cicero team,

I have a question about both version 2 and 3 of your lovely monocle. I am using cicero to create a developmental trajectory from atac-seq data. This works fine, but I want to next create monocle CDS objects containing either TSS activity or chromVAR motif enrichment with my own normalizations, and subsequently I want to transfer the mapping and graph info from the initial atac-seq peaks CDS object to those two, so that I can visualize the dynamics of TSS activity or motif enrichemnt on the same graph. N.B. I tried to just steal the following metadata from the initial object, but I am not 100% sure if its the best way to do it. What do you think? new_cds@reducedDimS <- original_cds@reducedDimS
new_cds@cellPairwiseDistances <- original_cds@cellPairwiseDistances
new_cds@reducedDimA <- original_cds@reducedDimA new_cds@minSpanningTree <- original_cds@minSpanningTree new_cds@dim_reduce_type <- original_cds@dim_reduce_type

Thank you!

hpliner commented 3 years ago

Hi, what happens when you attempt to run it/continue your analysis? Off the top of my head I think that might work - I believe you're replacing everything necessary, but depending on which downstream steps you are hoping to use, they may or may not work.

aleksandrPi commented 3 years ago

It seems to work well, only thing that is not getting transferred is the branch-points from the old trajectory, I could not find where they are contained...

hpliner commented 3 years ago

The branch points are stored in cds@auxOrderingData, might try transferring that as well

hpliner commented 3 years ago

Going to close, let me know if you have further issues