Retention time prediction workflow

Whole retention time prediction workflow. Function creates predictor set with RCDK based on SMILES. Trains a chosen predcition model and validates the approach with a cross validation.

Usage

fastret.workflow(
  data,
  method = "glmnet",
  verbose = FALSE,
  data_set_name = "data set",
  final_model = TRUE,
  preprocessed = FALSE,
  interaction_terms = FALSE,
  nfolds = 2,
  include_polynomial = FALSE,
  degree_polynomial = 2,
  scale = TRUE
)

Arguments

data

data.frame with columns NAME, RT, SMILES

method

prediction algorithm, either glmnet or xgboost

verbose

additional print outputs to user if TRUE

data_set_name

name of dataset will appear on validation plot

final_model

TRUE if final model trained on whole dataset should be returned

preprocessed

TRUE if data is already preprocessed and descriptor varialbes are already added

interaction_terms

TRUE if interaction terms between all variables should be added

nfolds

number of folds for cross validation

include_polynomial

TRUE if polynomial terms should be added to descriptor set

degree_polynomial

specifies degree up until which polynomials will be added if include_polynomials == TRUE

scale

if TRUE, all variables will be centered to a mean of 0 and scaled to a standard deviation of 1