Run the PRIMEmodel focal Pipeline on CTSS and Identified Regions

This function executes a partial PRIMEmodel pipeline, which includes the steps of validating and extending existing regions— such as those predicted from pooled PRIMEmodel data— followed by profile generation, model prediction, and BED result import.

Usage

predictFocal(
  ctss_rse,
  tc_gr,
  python_path = NULL,
  num_cores = NULL,
  keep_tmp = FALSE,
  log_dir = NULL
)

Arguments

ctss_rse

A `RangedSummarizedExperiment` object representing CTSS data.

tc_gr

A `GRanges` object representing identified regions. Regions will be extended to 401 bp width if needed.

python_path

Character path to the Python binary in the desired environment. Default is NULL.

num_cores

Optional integer specifying the number of CPU cores to use for parallel steps.

keep_tmp

Logical. If `TRUE`, temporary files and folders will be retained. Default is `FALSE`.

log_dir

Optional path to save a log file. If `NULL`, logs are printed to the console.

Value

A `GRanges` object if one sample was processed, or a `GRangesList` object if multiple samples were processed.

Details

This function supports downstream analysis by: - Validating and, if necessary, extending input regions to 401 bp - Generating normalized CTSS-based profiles per sample - Running predictions using a pre-trained PRIME model via Python - Importing the prediction results from BED files

Temporary files are stored in a subdirectory of `tempdir()` and removed unless `keep_tmp = TRUE`.