RNA velocity estimation in R
velocyto.R can be installed on unix-flavored systems, and requires the following key elements:
- C++11
- Open MP support
- boost libaries
- igraph library
- hdf5c++ library (as required by the h5 R package to support loom files)
The easiest way to install velocyto.R is using devtools::install_github() from R:
library(devtools)
install_github("velocyto-team/velocyto.R")
You need to have boost (e.g. sudo apt-get install libboost-dev) and openmp libraries installed. You can see detailed installation commands in the dockers/debian9/Dockerfile.
If you are having trouble installing the package on your system, you can build a docker instance that can be used on a wide range of systems and cloud environments. To install docker framework on your system see installation instruction. After installing the docker system, use the following commands to build a velocyto.R docker instance:
cd velocyto.R/dockers/debian9
docker build -t velocyto .
docker run --name velocyto -it velocytoThe example shows how to annotate SMART-seq2 reads from bam file and estimate RNA velocity.
The example shows how to load spliced/unspliced matrices from loom files prepared by velocyto.py CLI, use pagoda2 to cluster/embed cells, and then visualize RNA velocity on that embedding.
This example shows how to start analysis using dropEst count matrices, which can calculated from inDrop or 10x bam files using dropEst pipeline. It then uses pagoda2 to cluster/embed cells, and then visualize RNA velocity on that embedding.