expand_clusters: Expand clusters with unclustered isoforms
In ConesaLab/acorde: Isoform co-usage networks from single-cell RNA-seq data

expand_clusters

R Documentation

Expand clusters with unclustered isoforms

Description

Assign unclustered isoforms based on their correlation with the average expression profile of previously-generated clusters.

Usage

expand_clusters(
  data,
  isoform_col = NULL,
  id_table,
  cluster_list,
  unclustered,
  percentile_no = 10,
  force_expand = TRUE,
  expand_threshold = NULL,
  allow_negative_cors = TRUE,
  method = c("percentile", "pearson", "spearman", "rho", "zi_kendall")
)

Arguments

`data`	A data.frame or tibble object including isoforms as rows and cells as columns. Isoform IDs can be included as row names (data.frame) or as an additional column (tibble).
`isoform_col`	When a tibble is provided in `data`, a character value indicating the name of the column in which isoform IDs are specified.
`id_table`	A data frame including two columns named `cell` and `cell_type`, in which correspondence between cell ID and cell type should be provided. The number of rows should be equal to the total number of cell columns in `data`, and the order of the `cell` column should match column (i.e. cell) order in `data`.
`cluster_list`	A list of character vectors, each containing the identifiers of the isoforms in a cluster.
`unclustered`	A character vector containing the identifiers of unclustered isoforms.
`percentile_no`	Integer indicating the number of percentiles that will be used to summarized cell type expression via `percentile_expr`. Should always be higher than 4 (quantiles) and lower than 100 (percentiles). Defaults to 10.
`force_expand`	Logical. When `TRUE`, forced expansion is enabled, and unclustered isoforms are assigned to clusters based on the maximum reported co-expression. If `FALSE`, a co-expression threshold needs to be specified via `expand_threshold`.
`expand_threshold`	A numeric value defining the minimum correlation required to assign an unclustered isoform to a cluster.
`method`	Character indicating a co-expression method to use for merging similar clusters. Should be one of `percentile, pearson, spearman, zi_kendall, rho` (see details). Percentile correlation is used by default.
`allow_negative_cors.`	Logical. If set to `FALSE` (default), negative correlations will not be considered. Defaults to `TRUE`.

Details

Correlation-based cluster expansion first requires cluster metatranscripts to be calculated. A cluster's metatranscript is calculated as the mean of the percentile-summarized expression of all of the isoforms in that cluster. Next, co-expression values between the metatranscripts and the unclustered isoforms are computed using the similarity metric specified in method.

Available co-expression metrics (selected via the method) include:

percentile: percentile correlations computed using percentile_cor.
pearson: Pearson correlation computed using cor.
spearman: Spearman correlation computed using cor.
zi_kendall: zero-inflated Kendall correlation computed using the dismay function.
rho: rho proportionality metric computed using the dismay function.

Cluster expansion can be forced or threshold-based, depending on the values provided to force_expansion (i.e. TRUE or FALSE, respectively). If forced, isoforms will be assigned to a cluster if its metatranscript is the one yielding the highest co-expression with the isoform's expression, regardless of the exact co-expression metric value. Conversely, when threshold-based, only isoforms showing co-expression above the user-defined threshold (expand_threshold) with at least one metatranscript will be assigned to clusters. In this case, the maximally- correlated cluster is selected as the best match if there are several candidate clusters with above-threshold co-expression values.

Value

A list containing expanded clusters, where each element will include the identifiers of the isoforms assigned ot each cluster. If force_expand = FALSE, the first element of the list will contain the identifiers of isoforms that remained unassigned.