leapR
leapR.RdleapR is a wrapper function that consolidates multiple enrichment methods.
Arguments
- geneset
is a list of four vectors, gene names, gene descriptions, gene sizes and a matrix of genes. It represents .gmt format pathway files.
- enrichment_method
is a character string specifying the method of enrichment to be performed, one of: "enrichment_comparison", "enrichment_in_order", "enrichment_in_sets", "enrichment_in_pathway", "correlation_enrichment".
- eset
is a `SummarizedExperiment` object containing expression data, with features as rows and n sample/conditions as columns.
- assay_name
is the assay to be analyzed within the `eset`. Recommended to describe the data type (e.g. transcriptomics, proteomics) so that it can be integrated in `combine_omics`
- ...
further arguments
Details
Further arguments and enrichment method optional argument information:
| id_column | Is a character string, present in the rowData slot,
that is used to specify a column for identifiers to map to enrichment
libraries.
If missing, the rownames of the SummarizedExperiment assay will be used. |
| primary_columns | Is a character vector composed of column names from
eset (either in the `assay` or in the `rowData`),
that specifies a set of primary columns to calculate enrichment on.
The meaning of this varies according to the enrichment method used - see
the descriptions for each method below.
This is an optional argument used with 'enrichment_in_order',
'enrichment_in_sets', and 'enrichment_comparison' methods. |
| secondary_columns | |
| Is a character vector of column names for comparison, pulled from the `assay` of the SummarizedExperiment. This is an optional argument used with 'enrichment_comparison' methods. | |
| threshold | Is a numeric value, an optional argument used with 'enrichment_in sets' method which filters out abundance values or p-values (depending on what `primary_columns` is used) either above or below it. |
| greaterthan | |
Is a logical value that defaults to TRUE, it's used with
'enrichment_in_sets' method.
When set to TRUE, genes with `primary_columns` value above the
threshold argument are kept.
When set to FALSE genes with `primary_columns` value below the
threshold argument are kept.
This is an optional argument used with 'enrichment_in_sets' method. | |
| minsize | Is a numeric value, an optional argument used with 'enrichment_in_sets' and 'enrichment_in_order". |
| fdr | |
| A numerical value which specifies how many times to randomly sample genes to calculate an empirical false discovery rate, is an optional argument used with 'enrichment_comparison' method. | |
| min_p_threshold | Is a numeric value, a lower p-value threshold and is an optional argument used with 'enrichment_comparison' method. |
| sample_n | |
| Is a way to subsample the number of components considered for each calculation randomly. This is an optional argument used with 'enrichment_comparison' method. |
Enrichment Methods:
enrichment_comparison
Compares the distribution of abundances between two sets of
conditions for each pathway using a t test. For each pathway in
geneset uses a t test to compare the distribution of abundance
values/numbers in eset primary_columns with those in
eset secondary_columns. Lower p-values for pathways indicate
that the expression of the pathway is significantly different between the
set of conditions in primary_columns and the set of conditions in
secondary_columns.
Optionally, users can specify fdr which will calculate an empirical
p-value by randomizing abundances fdr number of times. If the
min_p_threshold is specified the method will only return pathways
with an adjusted p-value lower than the specified threshold. If
sample_n is specified the method will subsample the
pathway members to the specified number of components.
enrichment_in_order
Calculates enrichment of pathways based on a ranked list using the
Kolmogorov-Smirnov test. For each pathway in geneset uses a
Kolmogorov-Smirnov test for rank order to test if the distribution of ranked
abundance values in the eset primary_columns is significant
relative to a random distribution. Note that currently
primary_columns only accepts a single column for this method.
enrichment_in_sets
Calculates enrichment in pathway membership in a list (e.g. highly
differential proteins) relative to background using Fisher's exact test. For
each pathway in geneset uses a Fisher's exact test over- or under-
representation of a list of components specified. If targets are
specified this must be a vector of identifiers to serve as the target list
for comparison. If eset and primary_columns are specified then
threshold specifies a threshold value for determining the target list
of components to test. Specifying greaterthan to be False
will result in components with values lower than the specified
threshold. If eset is a data frame or matrix, the background
used for calculation will be taken as the rownames of eset
enrichment_in_pathway
Compares the distribution of abundances in a pathway with the background
distribution of abundances using a t test. For each pathway in
geneset calculates the significance of the difference between the
abundances from pathway members versus abundance of non-pathway members in
the set of conditions specified by primary_columns. Optionally, users
can specify fdr which will calculate an empirical p-value by
randomizing abundances fdr number of times. If the
min_p_threshold is specified the method will only return pathways
with an adjusted p-value lower than the specified threshold. If
sample_n is specified the method will subsample the
pathway members to the specified number of components.
correlation_enrichment
Calculates the enrichment of a pathway based on correlation between pathway
members across conditions versus correlation between members not in the
pathway. For each pathway in geneset calculates the pairwise
correlation between all pathway members and non-pathway members
across the specified primary_columns conditions in eset. Note
that for large matrices this can take a long time. A p-value is calculated
based on comparing the correlation within the members of a pathway with the
correlation values between members of the pathway and non-members of the
pathway.
Examples
library(leapR)
# read in the example abundance data
# read in the example transcriptomic data
tdata <- download.file("https://api.figshare.com/v2/file/download/56536214",
method='libcurl',destfile='transData.rda')
load('transData.rda')
p <- file.remove("transData.rda")
# read in the pathways
data("ncipid")
# read in the patient groups
data("shortlist")
data("longlist")
# use enrichment_comparison to calculate enrichment in one set of
# conditions (shortlist) and another (longlist)
short_v_long = leapR(geneset=ncipid, assay_name='transcriptomics',
enrichment_method='enrichment_comparison',
eset=tset, primary_columns=shortlist,
secondary_columns=longlist)
# use enrichment_in_sets to calculate the most enriched pathways
# from the highest abundance proteins
# from one condition
onept_sets = leapR(geneset=ncipid, assay_name='transcriptomics',
enrichment_method='enrichment_in_sets',
eset=tset, primary_columns="TCGA-13-1484", threshold=1.5)
# use enrichment_in_order to calculate the most enriched pathways from the
# same condition
# Note: that this uses the entire set of abundance values and their order -
# whereas the previous example uses a hard threshold to get a short list of
# most abundant proteins and calculates enrichment based on set overlap.
# The results are likely to be similar - but with some notable differences.
onept_order = leapR(geneset=ncipid, assay_name='transcriptomics',
enrichment_method='enrichment_in_order',
eset=tset, primary_columns="TCGA-13-1484")
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
#> Warning: p-value will be approximate in the presence of ties
# use enrichment_in_pathway to calculate the most enriched pathways in a
# set of conditions based on abundance in the pathway members versus
# abundance in non-pathway members
short_pathways = leapR(geneset=ncipid, assay_name='transcriptomics',
enrichment_method='enrichment_in_pathway',
eset=tset, primary_columns=shortlist)
# use correlation_enrichment to calculate the most enriched pathways in
# correlation across the shortlist conditions
short_correlation_pathways = leapR(geneset=ncipid,
assay_name='transcriptomics',
enrichment_method='correlation_enrichment',
eset=tset, primary_columns=shortlist)