run_carniphal: Run CARNIPHAL

In saezlab/CARNIPHAL: tool to run CARNIVAL on phosphoproteomic data

Description

This function runs CARNIVAL with the input of phosphoproteomic data (phosphosites and kinases). The prior knowledge network used is the combination of protein-protein and protein-phosphosite interactions from omnipath. Before running CARNIVAL the network is pruned by removing nodes n_steps upstream and downstream of measurements and inputs, respectively.

Usage

run_carniphal(
  inputObj,
  measObj,
  netObj = carniphalPKN,
  rmNodes = NULL,
  pruning = TRUE,
  n_steps_pruning = 50,
  solverPath,
  solver = "cplex",
  timelimit = 7200,
  mipGAP = 0.05,
  poolrelGAP = 0
)

Arguments

inputObj	named vector of perturbation targets. Either 1 (up regulated) or -1 (down regulated)
measObj	named vector of the measurements
netObj	data frame of the prior knowledge network
rmNodes	character vector of nodes to remove from prior knowledge network
pruning	logic, set to TRUE if network should be pruned (recommended)
n_steps_pruning	integer giving the order of the neighborhood
solverPath	path to the solver
solver	one of the solvers available from getSupportedSolvers()
timelimit	solver time limit in seconds
mipGAP	CPLEX parameter: absolute tolerance on the gap
poolrelGAP	CPLEX/Cbc parameter: Allowed relative gap of accepted

Value

List of CARNIVAL results and final inputObj, measObj, netObj used

saezlab/CARNIPHAL documentation built on Dec. 22, 2021, 9:12 p.m.