read.gene.mapping.info: read in detailed molecular mapping info from hdf5 file as...
In sqjin/nlvelo: RNA velocity estimation using nonlinear models

read in detailed molecular mapping info from hdf5 file as written out by "-d" option of velocyto.py

read.gene.mapping.info(
  fname,
  cell.clusters = NULL,
  internal.priming.info = NULL,
  min.exon.count = 10,
  n.cores = defaultNCores(),
  engine = "hdf5r"
)

`fname`	name of the hdf5 detailed molecular mapping (debug) file, written out by velocyto.py
`cell.clusters`	optional cell cluster factor
`internal.priming.info`	optionall internal priming info, as produced by find.ip.sites() function
`min.exon.count`	minimal total (dataset-wide) number of molecules for an exon to be considered expressed
`n.cores`	number of cores to use
`engine`	use either h5 or hdf5r to read the input hdf5 file

a list containing gene structural information data structure ($gene.df, with el,il, nex,nipconc,nipdisc columns corresponding to the log10 exonic length, intronic length, number of exons, numebr of internal concordant and discordant priming sites, respectively), and $info tables from the hdf5 file with an additional per-cluster entry $cluster.feature.counts table showing per-feature (rows) per-cluster (column) molecule counts (if cell.clusters are not supplied $info$cluster.feauture.counts will contain one column, 'all' giving dataset-wide counts)

sqjin/nlvelo documentation built on Feb. 11, 2021, 8:09 a.m.