Corrects the classification of reads contained in the class_data table.

Introduces two additional columns (corr_class and corr_comments) to the class_data table, reclassifying reads based on the positional quality flags computed by correct_residue_data. Reads whose all non-A positions were flagged as ambiguous (qc_pos == "N") are downgraded from "decorated" to "blank".

Usage

correct_class_data(residue_data_edited, class_data)

Arguments

residue_data_edited

Data frame or tibble. Corrected non-A residue predictions produced by correct_residue_data. Must contain columns mode_pos, seg_err_quart, and qc_pos.

class_data

Data frame or tibble containing read_classes predictions from the ninetails pipeline.

Value

A tibble identical to class_data with two additional columns:

corr_class

Character. Corrected classification: original class value, or "blank" for artefact-only reads.

corr_comments

Character. Corrected comment: original comments value, or "MPU" for reclassified reads.

Details

The reclassification logic: for each read, if every non-A residue position has qc_pos == "N", the read is reclassified as "blank" with comment "MPU" (to maintain compatibility with the tag system used in plotting functions). Reads with at least one qc_pos == "Y" position retain their original class and comment.

Caution

It is recommended to use reclassify_ninetails_data for downstream analyses, as it wraps this function together with correct_residue_data and performs the necessary column renaming. If using this function directly, rename corr_class and corr_comments to class and comments before plotting.

See also

correct_residue_data for the preceding step, reclassify_ninetails_data for the high-level wrapper, create_outputs for the pipeline that produces the input data.

Examples

if (FALSE) { # \dontrun{

class_data_corrected <- ninetails::correct_class_data(
  residue_data_edited = residue_data_edited,
  class_data = results[[1]])

} # }