dg__fitness_from_model: Predict fitness from computed dG model

In jschmiedel/tempura: Fit free energies of folding and binding from doubledeepPCA data

Description

Predict fitness from computed dG model

Usage

dg__fitness_from_model(
  dg_model0,
  variant_data0,
  varlist,
  parlist,
  calc_performance = FALSE,
  per_testset = FALSE,
  overwrite = TRUE,
  predfit_ddgvar = FALSE
)

Arguments

dg_model0	dG model(s) for which fitness is to be predicted, either long or wide format
variant_data0	variant_data data.table
varlist	varlist
parlist	parlist
calc_performance	logical, if Pearson's R should be calculated for predicted versus measured fitness, default = FALSE
per_testset	logical, if TRUE, predicts fitness only for test set variants (dg_model0 needs to contain test_set variable and is expected to be in wide format), default = FALSE
overwrite	logical, overwrite predicted fitness values if already present in variant_data0?, default = TRUE
predfit_ddgvar	logical, predict fitness values for a set of variations of ddg values (setting one to zero, holding parameters equal etc, used for plotting analysis), default = FALSE

Value

returns a list with a variant_data data.table with predited fitness values, and a prediction_performance data.table if calc_performance == TRUE

jschmiedel/tempura documentation built on Nov. 13, 2020, 3:53 a.m.