Skip to contents

Function to convert Allen-formatted MERFISH data to WSPmm format

create_count_data_WSPmm_allen.Rd

This is the create_count_data_WSPmm counterpart for MERFISH data already in Allen Brain Atlas format, i.e. already in long form with one row per gene/cell pair (columns trscrpt_gene_symb, trscrpt_ct, cluster_id, mouse), rather than the wide, one-column-per-gene form that create_count_data_WSPmm expects. Because the data is already long, no gene.list argument is needed here: genes are read directly from trscrpt_gene_symb. An age fixed effect, if requested via fixed.effect.names, is always set to "p5x" for every row, since this function was written for a single age cohort; include "age" in fixed.effect.names only when that is the cohort you are working with.

Usage

create_count_data_WSPmm_allen(
  df.merfish,
  bin.dim = c("x_bins", "y_bins"),
  fixed.effect.names,
  context = NULL
)

Arguments

df.merfish

A data frame of Allen-formatted MERFISH data, in long form with one row per gene/cell pair

bin.dim

A character string specifying the dimension by which to bin the data; must be either "x_bins" or "y_bins"

fixed.effect.names

A character vector of column names in df.merfish to include as fixed effects in the count data

context

A character string specifying the context level for fixed effects; if NULL, will use "context" literally for every row; otherwise must be the name of an existing column in df.merfish, whose values are used as the context level

Value

A data frame in count data format suitable for input into the WSPmm model

Examples

# A tiny toy data frame in the long, one-row-per-gene/cell-pair Allen format.
toy_allen <- data.frame(
  mouse             = rep(1, 4),
  cluster_id        = 1:4,
  y_bins            = c(1, 1, 2, 2),
  trscrpt_gene_symb = c("geneA", "geneB", "geneA", "geneB"),
  trscrpt_ct        = c(3, 1, 4, 2),
  hemisphere        = c("left", "right", "left", "right")
)
create_count_data_WSPmm_allen(
  toy_allen,
  bin.dim            = "y_bins",
  fixed.effect.names = "hemisphere"
)
#>   count bin context  gene mouse cell_num hemisphere
#> 1     3   1 context geneA     1        1       left
#> 2     1   1 context geneB     1        2      right
#> 3     4   2 context geneA     1        3       left
#> 4     2   2 context geneB     1        4      right