Optimized Virtual Knockout Experiment

Performs in-silico gene knockout experiments using optimized algorithms

Usage

scTenifoldKnk(
  countMatrix,
  gKO = NULL,
  qc_mtThreshold = 0.1,
  qc_minLSize = 1000,
  nc_lambda = 0,
  nc_nNet = 10,
  nc_nCells = 500,
  nc_nComp = 3,
  nc_scaleScores = TRUE,
  nc_symmetric = FALSE,
  nc_q = 0.9,
  td_K = 3,
  td_maxIter = 1000,
  td_maxError = 1e-05,
  td_nDecimal = 3,
  ma_nDim = 2,
  nCores = NULL,
  verbose = TRUE
)

Arguments

countMatrix

Gene expression count matrix (genes x cells)

gKO

Gene(s) to knockout

qc_mtThreshold

Maximum mitochondrial read ratio

qc_minLSize

Minimum library size

nc_lambda

Lambda parameter for strict directionality

nc_nNet

Number of networks to generate

nc_nCells

Number of cells per network

nc_nComp

Number of principal components

nc_scaleScores

Whether to scale network scores

nc_symmetric

Whether to make network symmetric

nc_q

Quantile threshold for edge filtering

td_K

Number of tensor components

td_maxIter

Maximum tensor decomposition iterations

td_maxError

Tensor decomposition error tolerance

td_nDecimal

Number of decimal places

ma_nDim

Number of manifold dimensions

nCores

Number of cores for parallel processing

verbose

Whether to print progress information

Value

A list containing tensor networks, manifold alignment, and differential regulation results

Examples

# Loading single-cell data
scRNAseq <- system.file("single-cell/example.csv", package = "scTenifoldKnk")
scRNAseq <- read.csv(scRNAseq, row.names = 1)

# Running scTenifoldKnk
result <- scTenifoldKnk(countMatrix = scRNAseq, gKO = "G100", qc_minLSize = 0)
#> === scTenifoldKnk: Virtual Knockout Analysis ===
#> 
#> Step 1/6: Quality control...
#>   Retained 100 genes and 2837 cells after QC
#> 
#> Step 2/6: Constructing gene regulatory networks...
#> Using sequential processing (dataset size doesn't warrant parallel overhead)
#> Generating 10 networks (100 genes, 500 cells/network, 3 PCs)...
#> Building networks sequentially...
#>   Network 1/10
#>   Network 6/10
#> Successfully generated 10 networks
#> 
#> Step 3/6: Tensor decomposition...
#>   Tensor decomposition complete
#> 
#> Step 4/6: Performing virtual knockout...
#>   Gene 'G100' knocked out
#> 
#> Step 5/6: Manifold alignment...
#>   Manifold alignment complete
#> 
#> Step 6/6: Differential regulation analysis...
#>   Found 1 significantly affected genes (FDR < 0.05)
#> 
#> === Analysis complete ===