multi_screener: Generate enzyme screener file

In edpratt1/KINATESTID: A Pipeline for Kinase-Specific Artificial Peptide Design

Description

The function multi_screener takes as inputs 1) the kinase-specific preferred substrates file and 2) the protein-matched in silico control peptide library. For each enzyme, an optimal threshold will be calculated to separate the peptide scores of sample vs control sequences. This is calculated using the package 'cutpointr.' A bootstrap model is used to select an optimal cutpoint to maximize sensitivity.

Usage

multi_screener(
  screener_input,
  uniprot_input,
  path,
  method = c("prod", "log2_sum", "w_prod"),
  pval_corr = FALSE,
  type = "aa",
  norm_method = c("none", "bkgrnd"),
  property = NULL,
  constrain = NULL
)

Arguments

screener_input	An enzyme-specific preferred substrates file.
uniprot_input	The 'uniprot_id'-matched in silico control peptide library.
path	The file path where output data should be saved.
method	Peptide scoring algorithm. Accepts product of odds ratios ('prod'), weighted product of odds ratio ('w_prod'), and the sum of log scores ('log2_sum').
pval_corr	A logical parameter where TRUE will set all fisher odds with non-significant p-values (> 0.05) to 1. Set to FALSE by default.
type	Character of either 'aa' or 'aa_property' to indicate whether amino acid residues or amino acid properties should be used.
norm_method	Character of either 'none' or 'bkgrnd' to indicate whether raw scores or background-corrected ones should be used. Background-correction subtracts the mean and divides by the standard deviation of the negative control substrate scores.
property	Character indicating which amino acid property should be analyzed. Only used if type = 'aa'. See available property options by typing 'aa_classificiation'.
constrain	The minimum specificity required when calculating the threshold score for activity. By default set to 0.9.

Value

A three-element list containing: 1) A lot of score distributions and the selected threshold, 2) Fisher tables for each kinase in the 'screener_input' file, and 3) Summary stats on score distribution for each enzyme.

Examples

screener <- multi_screener(screener_raw, screener_uniprot, 
                           path = output_dir,
                           method = "prod",
                           pval_corr = FALSE,
                           type = "aa",
                           norm_method = "none",
                           constrain = 0.90)

edpratt1/KINATESTID documentation built on Feb. 5, 2022, 1:21 p.m.