SignatuR

An R package to store and interact with gene signatures for single-cell analysis

Install

Get the database and its accessor functions:

remotes::install_github("carmonalab/SignatuR")
library(SignatuR)
data(SignatuR)

Data format

The DB is structured as a data.tree object, with signatures organized by species and category. Gene lists are stored in the "Signature" attribute, comments and references for the signature are stored in the "Reference" attribute.

Several functions for easy interaction with the data structure have been implemented. See some examples below:

• See database structure

data(SignatuR)
SignatuR

• See database with annotations

print(SignatuR, "Reference","Signature")

• Plot database structure (requires installing DiagrammeR)

library(DiagrammeR)
plot(SignatuR)

• Extract a specific signature

s <- GetSignature(SignatuR$Mm$Programs$HeatShock)

• Extract all signatures below a given node

ss <- GetSignature(SignatuR$Mm$Programs)

• Add a new signature to the DB (e.g. to "Cell_types" node)

SignatuR <- AddSignature(SignatuR,
	node=SignatuR$Mm$Cell_types,
	name="T_cell",
	reference="A simple T cell signature",
	signature=c("Cd2","Cd3d","Cd3e"))

• Modify and existing signature (turn on 'overwrite' flag)

SignatuR <- AddSignature(SignatuR,
	node=SignatuR$Mm$Cell_types,
	name="T_cell",
	reference="A modified T cell signature",
	signature=c("Cd2","Cd3d","Cd3e","Cd3g"),
	overwrite=T)

• Add a new node to the DB

SignatuR <- AddNode(SignatuR,
  parent_node=SignatuR$Hs,
  name="New_category")

• Remove a signature or a node from the DB

RemoveSignature(SignatuR$Hs$Compartments$TCR)

• Make a clone of the DB (allows you to edit without breaking the DB)

SignatuR.copy <- data.tree::Clone(SignatuR)

• Save a local copy of your modified SignatuR DB

SaveSignatuR(SignatuR, file="mySignatuR.csv")

• Load a local copy of SignatuR from disk

mySignatuR <- LoadSignatuR("mySignatuR.csv")

• Save updated database (for package developers)

usethis::use_data(SignatuR, overwrite = TRUE)

Use examples

• In gene feature selection for data integration