ampir (version 1.1.0)

predict_amps: Predict the antimicrobial peptide probability of a protein

Description

This function predicts the probability of a protein to be an antimicrobial peptide

Usage

predict_amps(faa_df, min_len = 5, n_cores = 1, model = "precursor")

Arguments

faa_df

A dataframe obtained from read_faa containing two columns: the sequence name (seq_name) and amino acid sequence (seq_aa)

min_len

The minimum protein length for which predictions will be generated

n_cores

On multicore machines split the task across this many processors. This option does not work on Windows

model

Either a string with the name of a built-in model (mature, precursor), OR, A train object suitable for passing to the predict.train function in the caret package. If omitted the default model will be used.

Value

The original input data.frame with a new column added called prob_AMP with the probability of that sequence to be an antimicrobial peptide. Any sequences that are too short or which contain invalid amin acids will have NA in this column

Examples

# NOT RUN {
my_bat_faa_df <- read_faa(system.file("extdata/bat_protein.fasta", package = "ampir"))

predict_amps(my_bat_faa_df)
#       seq_name    prob_AMP
# [1] G1P6H5_MYOLU  0.9723796
# }