TF_Filter: Generate network

In lusystemsbio/NetAct: NetAct -- a computational platform to construct core transcription factor regulatory networks using gene activity

Generate network

Description

Network calculated using activity and interaction database. Uses mutual information to find possible interactions and keeps the interactions if they are available in the database. Sign of interaction is assigned based on the correlation between the activities.

Usage

TF_Filter(
  actMat,
  GSDB,
  miTh = 0.4,
  maxTf = 75,
  maxInteractions = 300,
  nbins = 16,
  miMethod = "2d",
  corMethod = "spearman",
  useCor = FALSE,
  removeSignalling = FALSE,
  DPI = FALSE,
  nameFile = NULL,
  ...
)

Arguments

actMat	numeric matrix. Matrix containing the activities
GSDB	List of list. Gene set database of interactions
miTh	numeric. Mutual information threshold
maxTf	integer (optional). Default 75. Maximum number of transcription factors in the network. If removeSignalling is TRUE the actual number will be less.
maxInteractions	integer (optional). Default 300. Maximum number of interactions in the network.
nbins	integer (optional). Number of bins Default 16.
miMethod	MI calculation method: "2d": 2D discretization with entropy (default) "1d": 1D discretization with infotheo
corMethod	character (optional). Method to compute correlation.
useCor	Logical (optional). Whether to use correlation instead of mutual information to find possible interactions.
removeSignalling	logical (optional). Whether to remove the Tfs which are not the target of any other Tfs. Default TRUE. It is not recursive and the generated network might still contain some signalling tfs.
DPI	logical (optional). Default FALSE. Whether to apply the data processing inequality to remove weak edges from triangles.
nameFile	character (optional). Ouput file name. Default NULL (no file output).
...	two additional parameters passed from applyDPI (default: miDiff = 0, minMiTh = 0.5)

Value

data.frame. Contains the interactions in a dataframe listing. source tf, target tf and interaction type (1-activation, 2-inhibition).

lusystemsbio/NetAct documentation built on Sept. 7, 2024, 8:34 p.m.