This function summarizes isoform-level PI values and visualizes the relationship between mean PI and Phi across isoforms. Isoforms can be filtered based on their mean PI and total gene-level support before plotting.
Arguments
- pi_mat
A numeric or sparse matrix of PI values (isoforms × cells).
- iso_mat
A numeric or sparse matrix of isoform-level counts (same dimensions and rownames as
pi_mat), used to derive gene-level coverage.- pi_min
Minimum mean PI threshold to retain isoforms (default: 0.3).
- pi_max
Maximum mean PI threshold to retain isoforms (default: 0.7).
- min_gene_count
Minimum total gene count required to include an isoform in the plot (default: 3500).
Details
For each isoform, the function computes:
mean PI across cells;
Phi using
computePHI(inter-cell variability);number of cells with non-zero counts;
total gene-level count (sum across all isoforms of a gene).
Isoforms are filtered according to pi_min, pi_max, and
min_gene_count, and only those passing all filters are plotted.
Examples
if (FALSE) { # \dontrun{
set.seed(123)
pi_mat <- matrix(runif(200, 0.05, 0.95), nrow = 50,
dimnames = list(paste0("Gene", rep(1:10, each = 5), "..Tx", 1:5),
paste0("Cell", 1:4)))
iso_mat <- matrix(sample(0:20, 200, replace = TRUE), nrow = 50,
dimnames = dimnames(pi_mat))
# Default thresholds
plotPIPHI(pi_mat, iso_mat)
# Custom thresholds
plotPIPHI(pi_mat, iso_mat, pi_min = 0.2, pi_max = 0.8, min_gene_count = 10)
} # }