doubledeepms__mochi__fit_tmodel_3state: doubledeepms__mochi__fit_tmodel_3state

In lehner-lab/doubledeepms: Analysis scripts to reproduce the figures and results from doubleDeepPCA manuscript

doubledeepms__mochi__fit_tmodel_3state

Description

Fit 3-state thermodynamic model to ddPCA data.

Usage

doubledeepms__mochi__fit_tmodel_3state(fitness_folding_inpath = NULL,
  fitness_binding_inpath = NULL, id_subset_folding_inpath = "all",
  id_subset_binding_inpath = "all", mut_order_subset_train = "1,2",
  mut_order_subset_valid = "2", mut_order_subset_obs = "union",
  min_multiplicity_valid = 2, fit_type = "fb",
  sequence_type = "aminoacid", subsample_prop = 1, mochi_script,
  mochi_outpath, num_epochs = 1000, num_epochs_grid = 100,
  validation_set_proportion = 0.1, num_resamplings = 10,
  num_samples = "128,256,512,1024",
  learning_rate = "0.0001,0.001,0.01,0.1", job_number = 1)

Arguments

fitness_folding_inpath	Folding fitness DiMSum RData file (default:NULL)
fitness_binding_inpath	Folding fitness DiMSum RData file (default:NULL)
id_subset_folding_inpath	Folding coefficient id subset file (default:all)
id_subset_binding_inpath	Folding coefficient id subset file (default:all)
mut_order_subset_train	Mutation order subset for training data, in addition to WT (default:1,2)
mut_order_subset_valid	Mutation order subset for validation data, in addition to WT (default:2)
mut_order_subset_obs	Mutation order subset for all observed data, in addition to WT (default:union of mutOrderSubsetTrain and mutOrderSubsetValid)
min_multiplicity_valid	Minimum multiplicity for validation data (default:2)
fit_type	Fit type, either fb=folding+binding, f=folding, b=binding (default:fb)
sequence_type	Sequence type (default:aminoacid)
subsample_prop	Doubles subsample proportion (default:1)
mochi_script	MoCHI script (required)
mochi_outpath	Output folder (required)
num_epochs	Number of epochs (default:1000)
num_epochs_grid	Number of epochs for gridsearch (default:100)
validation_set_proportion	Validation set proportion (default:0.1)
num_resamplings	Number of random resamplings of target values (default:10)
num_samples	Number of samples per gradient update (default:128,256,512,1024)
learning_rate	Learning rate (default:0.0001,0.001,0.01,0.1)
job_number	Job number (default:1)

Value

Nothing

lehner-lab/doubledeepms documentation built on July 21, 2023, 4:10 a.m.