ModelObj: ModelObj Environment

In CshlSiepelLab/ACE: Analysis of CRISPR Essentiality in R

Description

This R6 environment object processes the raw data loaded in the provided 'user_DataObj' to calculate model constants according to user specifications. Computes sample scaling parameters, the vector of infected cell values to use in the likelihood summation, and partitions test and control samples,

Value

ModelObj (R6Class). Contains: – mean_var_model_params = "list", – guide_features = "matrix", – unobserved_infected_cell_values = "vector", – mean_var_model = "integer", – init_scaling = "vector", – dep_scaling = "vector", – test_sample_subtype_cols = "vector" – stepSize = 10 – guidePrior = 'string'

Public fields

mean_var_model_params

Coefficients to use in mean-var model.

guide_features

Matrix of by-guide features.

master_freq_dt

Data.table of sgRNA frequencies in master libraries.

unobserved_infected_cell_values

Discrete values of infected cells for Riemann sum.

mean_var_model

Integer specifying which form of model to use.

init_scaling

By-sample vector of sample scaling factors.

dep_scaling

By-sample vector of depleted sample scaling.

test_sample_subtype_cols

Vector of column indices for 'test' subtype.

stepSize

Integer of infected cell intervals to evaluate.

guidePrior

String indicating form of infection prior for README.

use_neg_ctrl

Boolean indicating whether negative controls from DataObj should be used.

neg_ctrls

String vector of gene names for negative controls.

Methods

Public methods

• ModelObj$new()

• ModelObj$clone()

---

Method new()

Create ModelObj Write messages and warnings to log file. @param message_vector String or table to write.

Usage

ModelObj$new(
  user_DataObj = NA,
  use_master_library = T,
  fit_guide_parameter = F,
  mean_var_model = 1,
  use_neg_ctrl = T,
  test_samples = NA,
  stepSize = 10
)

Arguments

user_DataObj

Containing all count data; a DataObj.

use_master_library

Boolean, use masterlibrary to set guide abundance prior?

fit_guide_parameter

Boolean, should by-guide efficiencies be calculated?

mean_var_model

Optional integer for mean~var fitting model; default poisson.

use_neg_ctrl

Boolean, use negative control genes for scaling sample parameters and final essentiality value.

test_samples

String indicating which sample annotation to use when grouping test and control groups.

stepSize

Number of cells to use for each step in the summation, default 10.

---

Method clone()

The objects of this class are cloneable with this method.

Usage

ModelObj$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

CshlSiepelLab/ACE documentation built on Sept. 10, 2021, 11:21 p.m.