cluster_reactome_pathways: Cluster Reactome Pathways

In hancockinformatics/clusterPathways: Cluster Pathways Using Genes and Jaccard Index

Cluster Reactome Pathways

Description

Using a data frame of Reactome pathways, create a pairwise Jaccard matrix, then cluster pathways accordingly. Heatmaps/denrograms are saved, and two results tables are returned as output containing all the input pathways and which cluster they belonged to.

Usage

cluster_reactome_pathways(
  input_pathways,
  input_genes,
  species = "human",
  output_dir = NULL,
  width = 10,
  height = 20
)

Arguments

input_pathways	Table of Reactome pathways to be clustered. Must have the pathway ID in the first column, and description in the second. If a "direction" column (case sensitive) is present, it will be added to heatmap annotations (see Details for more information).
input_genes	Character vector of genes used to generate the input_pathways table, e.g. a list of DE genes. Must be Ensembl IDs.
species	Either "human" (the default) or "mouse".
output_dir	Directory to save heatmaps into. It will be created if it doesn't already exist.
width	Width and height of output heatmaps in inches.
height	Width and height of output heatmaps in inches.

Details

The direction column must contain either "up" or "down" for each pathway present. Both column name and contents are case-sensitive.

Value

A named list containing three data frames, "all_pathways", "rep_pathways", and "missing_pathways". (see below for more information). The function also saves heatmaps to ".png" files in the provided output directory.

The additional columns in the first two results tables are:

level_1, level_2

The highest and second-highest levels for each pathway from the Reactome hierarchy

n_bg_genes

The total number of genes annotated to the pathway

n_cd_genes

The number genes annotated to the pathway that were present in the provided input list (candidate genes)

gene_ratio

n_cd_genes / n_bg_genes

cluster

Denotes the group each pathway is placed within

The "rep_pathways" table contains all of the above, plus the column "n_pathways" which is the number of pathways from that cluster (the number in square brackets in the "heatmap_rep_pathways_clustered.png" output image). The "missing_pathways" table contains any input pathways that were not included in the analysis/figures, typically due to missing data.

References

None.

See Also

https://www.github.com/hancockinformatics/clusterPathways

Examples

## Not run: 
  library(clusterPathways)
  library(tidyverse)

  # Load table of pathways
  my_pathways <- read_csv("pathway_enrichment_result.csv")

  # Load list of genes, which were tested to obtain the above pathways
  my_genes <- read_csv("de_genes.csv") %>% pull(ensembl_gene_id)

  # Run cluster_reactome_pathways
  cluster_reactome_pathways(
    input_pathways = my_pathways,
    input_genes = my_genes,
    species = "human",
    output_dir = "clustered_pathways",
    width = 18,
    height = 30
  )

## End(Not run)

hancockinformatics/clusterPathways documentation built on March 19, 2022, 7:23 p.m.