fit: Fit model

In CshlSiepelLab/tuSelecter2: Deconvolution of Expression for Nascent RNA Sequencing Data

Description

Estimates trancript abundances for a given transcript_quantifier-class object under a lasso penalty

Usage

fit(
  tq,
  lambda = 0,
  transform = "log",
  inactive_transcripts = NA,
  verbose = FALSE,
  heuristic_inactive_override = T
)

## S4 method for signature 'transcript_quantifier'
fit(
  tq,
  lambda = 0,
  transform = "log",
  inactive_transcripts = NULL,
  verbose = FALSE,
  heuristic_inactive_override = T
)

Arguments

tq	A transcript_quantifier-class object
lambda	the weight of the lasso penalty (not implemented at this time)
transform	how to transform the data (either "identity" or "log")
inactive_transcripts	a character vector listing transcripts for which abundance values should be fixed to 0. IMPORTANT: In the case where multiple transcripts are assigned to a single model (due to identical models at the specified bin scale) this will be overridden if one or more transcripts assigned to the same model are active. This will also be overidden for specific transcripts if they have 10x or more polymerase density downstream from their TSS as they do upstream and there are no other active transcripts in their loci.
verbose	if TRUE shows progress bar for fitting (default: FALSE)
heuristic_inactive_override	Boolean. If TRUE uses a series of heuristics to re-activate some of the transcripts labeled as inactive. We reccomend you leave this on unless your're really confident in your inactive calls. See Details for more.

CshlSiepelLab/tuSelecter2 documentation built on July 18, 2021, 5:09 p.m.