scrattch.mapping

Generalized mapping scripts for RNA-seq, Patch-seq or any gene expression data. Assumes that you have built an Allen Institute Taxonomy (AIT) object with scrattch.taxonomy.

Workflow

Documentation

You can find a detail description of all scrattch.mapping functions here:

Update notes are here:

Docker

We have setup a docker environemnt for scattch.mapping that contains all the required dependencies and the current version of scrattch.mapping. This docker is accessible through docker hub via: njjai/scrattch_mapping:0.55.4.

HPC usage:

Non-interactive

singularity exec --cleanenv docker://njjai/scrattch_mapping:0.55.4 Rscript YOUR_CODE.R

Interactive

singularity shell --cleanenv docker://njjai/scrattch_mapping:0.55.4

Installation

While we advice using the provided docker, you can also install scrattch.mapping directly from github as follows:

Note: slight edits to installation will be needed while repo is private. Also note that doMC may need to be installed manually from the download at https://r-forge.r-project.org/R/?group_id=947 if you use Windows.

# Quickly, but without the vignettes:
devtools::install_github("AllenInstitute/scrattch-mapping")

# More slowly, but with the vignettes:
devtools::install_github("AllenInstitute/scrattch-mapping",build_vignettes=TRUE, force=TRUE)

Note that this strategy might not work outside the docker due to complicated dependencies. Vignettes are provided below.

Usage examples

1. Run Flat, Tree, and Seurat taxonomy mapping This examples shows how to use scrattch.mapping for standard taxonomy mapping.

Reporting issues

If you run into any issues, please let Nelson and Jeremy know or create a new issue in the 'Issues' tab above.

TODO

• Generalize HANN mapping for external use.
• Standardize output from multiple modes of mapping (RNA, Patch-Seq, spatial?)

Done

• Allow other values than 100 for bootstrapping
• Minor vignette updates (e.g, buildReference --> buildTaxonomy)
• Allow hGenes to impact corr and Seurat mapping, but have tree and HANN ignore it (it will break the node by node marker approach)
• Pass a dendrogram to buildTaxonomy if one exists already (I don't remember if this is allowed, but if not, it may be more complicated than it sounds, which is probably why I haven't done it).
• load the AIT.anndata directly from the .h5ad if it exists
• Fixes to addDendrogramMarker based on ET feedback.
• Package up the stand-alone script.
• Add stand-alone scripts for creating a reference sn/scRNA-seq taxonomy to package
• Accept various inputs (Seurat, Anndata, count matrix + metadata table, reading from reference taxonomy folder)?
• Write in a module for making the taxonomy compatible with tree mapping (adding markers to dendrogram). Perhaps this should be done after the taxonomy is made for generalization. (JM: I was planning to take this one)
• Add stand-alone scripts for patch-seq QC to package
• Output feather with gene counts along with required Shiny files for use with scVI/scANVI etc. I’ve had to manually make a few of these for the older BG taxonomies.
• This might be the same thing, but also output files for viewing data in existing patch-seq shiny folder format.
• Implement new tree mapping (Code from CK)
• Dockerize.