percent_spliced_in: Relative measure of splice variant usage, limited currently...

In snapcount: R/Bioconductor Package for interfacing with Snaptron for rapid querying of expression counts

Description

Similar to the JIR, this calculates Percent Spliced In (PSI) statistics for the definition of 2 different groups: inclusion and exclusion. Currently this function only supports the cassette exon use case.

Usage

percent_spliced_in(
  inclusion_group1,
  inclusion_group2,
  exclusion_group,
  min_count = 20,
  group_names = NULL
)

Arguments

inclusion_group1, inclusion_group2, exclusion_group	Where each is a list of 1 or more QueryBuilder objects
min_count	minimum total count (denominator) required to not be assigned -1
group_names	Optional vector of strings representing the group names

Details

Inclusion typically defines 2 basic queries, one for the junction preceding the cassette exon, and the second for the junction following the cassette exon. The exclusion group contains one basic query which defines the junction which skips the cassette exon.

The PSI itself is implemented as:

PSI(inclusion1, inclusion2, exclusion) = mean(inclusion1, inclusion2) / (mean(inclusion1, inclusion2) + exclusion)

where each term denotes the coverage of junctions that resulted from the basic queries in that group in the current sample.

Value

A DataFrame of samples, with their PSI score and metadata, which had > 0 coverage in at least one resulting row in at least one of the groups

Examples

in1 <- QueryBuilder(compilation = "srav2", regions = "chr1:94468008-94472172")
in1 <- set_coordinate_modifier(in1, Coordinates$Exact)
in1 <- set_row_filters(in1, strand == "+")

in2 <- QueryBuilder(compilation = "srav2", regions = "chr1:94468008-94472172")
in2 <- set_coordinate_modifier(in2, Coordinates$Exact)
in2 <- set_row_filters(in2, strand == "+")

ex <- QueryBuilder(compilation = "srav2", regions = "chr1:94468008-94475142")
ex <- set_coordinate_modifier(ex, Coordinates$Exact)
ex <- set_row_filters(ex, strand == "+")

percent_spliced_in(list(in1), list(in2), list(ex))

snapcount documentation built on Nov. 8, 2020, 4:53 p.m.