get_unsupervised_graph_from_files: Builds an unsupervised graph starting from a list of input...
In ParkerICI/scgraphs: Analyzing single-cell data using graphs

This function is similar to get_unsupervised_graph, except that the input in this case is a list of file names containing the vertices of the graph. Each file should be a tab-separated table of nodes, similar to the tab input of get_unsupervised_graph. This function will rbind all the files in files.list, and a build a single graph with the union of all the nodes. Optionally, file-level metadata can also be incorporated in the vertex properties of the resulting graph, using the metadata.tab parameter

get_unsupervised_graph_from_files(
  files.list,
  col.names,
  filtering.threshold,
  metadata.tab = NULL,
  metadata.filename.col = NULL,
  use.basename = TRUE,
  process.clusters.data = TRUE,
  clusters.data.out.dir = "./",
  downsample.to = 1000,
  method = c("forceatlas2", "umap"),
  ...
)

`files.list`	The list of files to process. The function will first determine the set of columns that are common to all the files. Only the common vertex properties will feature in the output graph
`col.names`	A character vector indicating which columns of `tab` should be used to calculate distances
`filtering.threshold`	The threshold used to filter edges in the graph
`metadata.tab`	Optional. If specified, a table of file-level metadata, to be added as vertex properties in the graph. Each row should specify metadata for a single file, with the columns of `metadata.tab` representing metadata values. All the vertices derived from that file will have the corresponding metadatata value. Please note that the names `name`, `Label`, `type` and `sample` are used internally by this package, and therfore cannot bs used as metadata vertex properties
`metadata.filename.col`	The name of the column in `metadata.tab` that contains the file name to be matched to the files in `files.list`
`use.basename`	The resulting graph will contain an additional vertex property called `sample` identifying which file the vertex was derived from. If `use.basename` is `TRUE` the `basename` will be used, otherwise the full path as specified in `files.list`. Moreover if `use.basename` is `TRUE` the matching with the file names contained in `metadata.tab` will be based on the `basename` only
`process.clusters.data`	If this is `TRUE` this function will look for a file with extension `.all_events.rds` for each file in `files.list` (see the Documentation of `grappolo::cluster_fcs_files`). This file contains single-cell data (i.e. each row represent a cell, as opposed to the files in `files.list` where each row represents a cluster). Each file will be processed using the `write_clusters_data` function. This processing is used for downstream data visualization but it is not strictly necessary to create the graph. If `use.basename` is `TRUE` the `basename` of the files in `files.list` will be used for processing.
`clusters.data.out.dir`	Only used if `process.clusters.data == TRUE`. The output directory where the clusters data will be written
`downsample.to`	The target number of events for downsampling. Only used if `process.clusters.data == TRUE`. This is only used for downstream data visualization and does not affect the construction of the graph
`method`	The method to use. Either build a force-directed layout graph using ForceAtlas2, or alternatively use UMAP
`...`	Additional argument passed to build_graph or build_umap_graph depending on the choice of method