VDJ_network: Similarity networks based on CDR3 regions

In alexyermanos/Platypus: Single-Cell Immune Repertoire and Gene Expression Analysis

Similarity networks based on CDR3 regions

Description

Creates a similarity network where clones with similar CDR3s are connected.

Usage

VDJ_network(
  VDJ,
  distance.cutoff,
  per.sample,
  platypus.version,
  known.binders,
  hcdr3.only,
  is.bulk
)

Arguments

VDJ	Either (for platypus version "v2") output from VDJ_analyze function. This should be a list of clonotype dataframes, with each list element corresponding to a single VDJ repertoire, OR (for platypus version "v3") the the VDJ matrix output of the VDJ_GEX_matrix() function (VDJ.GEX.matrix.output[[1]])
distance.cutoff	The threshold Levenshtein distance for which two nodes will be connected on the similarity network.
per.sample	logical value indicating if a single networks should be produced for each mouse.
platypus.version	Character. Defaults to "v3". Can be "v2" or "v3" dependent on the input format
known.binders	Either a character vector with cdr3s of known binders or a data frame with cdr3s in the first and the corresponding specificity in the second column. If this parameter is defined, the output will be a network with only edges between known binders and the repertoire nodes and edges between the known binders that have at least one edge to a repertoire node
hcdr3.only	logical value indicating if the network is based on heavy chain cdr3s (hcdr3.only = T) or pasted heavy and light chain cdr3s (hcdr3.only = F), works for platypus.version 3 only
is.bulk	logical value indicating whether the VDJ input was generated from bulk-sequencing data using the bulk_to_vgm function. If is.bulk = T, the VDJ_network function is compatible for use with bulk data. Defaults to False (F).

Value

returns a list containing networks and network information. If per.sample is set to TRUE then the result will be a network for each repertoire. If per.sample ==FALSE, output[[1]] <- will contain the network, output[[2]] will contain the dataframe with information on each node, such as frequency, mouse origin etc. output[[3]] will contain the connected index - these numbers indicate that the nodes are connected to at least one other node. output[[4]] contains the paired graph - so the graph where only the connected nodes are drawn.

Examples

try({
network_out <- VDJ_network(VDJ = Platypus::small_vgm[[1]],per.sample = FALSE,distance.cutoff = 2)
})

alexyermanos/Platypus documentation built on Feb. 8, 2025, 1:22 p.m.