This function creates an average metafeat/metagene/module for clusters.
Usage
createMetafeats(
gobject,
spat_unit = NULL,
feat_type = NULL,
expression_values = c("normalized", "scaled", "custom"),
feat_clusters,
stat = c("mean", "sum", "max", "min"),
rescale_to = NULL,
name = paste0("metafeat_", ifelse(is.function(stat), "custom", stat)),
return_gobject = TRUE,
verbose = NULL
)Arguments
- gobject
Giotto object
- spat_unit
spatial unit
- feat_type
feature type
- expression_values
expression values to use
- feat_clusters
numerical vector with features as names
- stat
statistical transformation to calculate across the cluster features per cell to get the metafeature value. Provided defaults are "mean", 'sum', "max", and "min". If a function is supplied instead, it will be run via
apply(MARGIN = 2)across the expression matrix to calculate the metafeature value.- rescale_to
numeric vector of length 2. (optional) New maximum and minimum to rescale all features to before calculating the metafeature score using
stat. Values are first zeroed by setting the minimum value to 0, then they are rescaled. In cases where all values are the same across a feature, the rescale produces the middle point of the range specified byrescale_to- name
name of the metagene results
- return_gobject
return giotto object
- verbose
be verbose
Details
An example for the 'gene_clusters' could be like this: cluster_vector = c(1, 1, 2, 2) names(cluster_vector) = c('geneA', 'geneB', 'geneC', 'geneD')
Examples
# load a dataset
g <- GiottoData::loadGiottoMini("viz")
#> 1. read Giotto object
#> 2. read Giotto feature information
#> 3. read Giotto spatial information
#> 3.1 read Giotto spatial shape information
#> 3.2 read Giotto spatial centroid information
#> 3.3 read Giotto spatial overlap information
#> 4. read Giotto image information
#> python already initialized in this session
#> active environment : 'giotto_env'
#> python version : 3.10
# set a spat unit to use
activeSpatUnit(g) <- "aggregate"
# define the metafeats
# We do this by making a set of annotations. These annotations can be
# made as either a vector or data.frame
# 1.1 annotation vector (numeric)
feats_to_use <- featIDs(g)[seq_len(6)]
clust_to_use1 <- c(1, 1, 1, 2, 2, 2)
names(clust_to_use1) <- feats_to_use
force(clust_to_use1)
#> Mlc1 Gprc5b Gfap Ednrb Sox9 Aqp4
#> 1 1 1 2 2 2
g <- createMetafeats(
gobject = g,
feat_clusters = clust_to_use1,
name = "metagene_1"
)
#> calculating metafeature: 1
#> calculating metafeature: 2
# 1.2 annotation vector (character)
clust_to_use2 <- c(rep("sig_a", 3), rep("sig_b", 3))
names(clust_to_use2) <- feats_to_use
force(clust_to_use2)
#> Mlc1 Gprc5b Gfap Ednrb Sox9 Aqp4
#> "sig_a" "sig_a" "sig_a" "sig_b" "sig_b" "sig_b"
g <- createMetafeats(
gobject = g,
feat_clusters = clust_to_use2,
name = "metagene_2"
)
#> calculating metafeature: sig_a
#> calculating metafeature: sig_b
# 2.1 annotation data.frame (no weights)
# Cols must be:
# `clus` - which metafeature
# `feat` - features to assign to this metafeature
# `w` - (optional) weight to assign the feature before metafeature score
# calculation
clust_to_use3 <- data.frame(
clus = c(rep("sig_a", 3), rep("sig_b", 3)),
feat = feats_to_use
)
force(clust_to_use3)
#> clus feat
#> 1 sig_a Mlc1
#> 2 sig_a Gprc5b
#> 3 sig_a Gfap
#> 4 sig_b Ednrb
#> 5 sig_b Sox9
#> 6 sig_b Aqp4
g <- createMetafeats(
gobject = g,
feat_clusters = clust_to_use3,
name = "metagene_3"
)
#> calculating metafeature: sig_a
#> calculating metafeature: sig_b
# 2.1 annotation data.frame (weights)
clust_to_use4 <- data.frame(
clus = c(rep("sig_a", 3), rep("sig_b", 3)),
feat = feats_to_use,
w = c(1, 1, 1, 2, 3, 4)
)
force(clust_to_use4)
#> clus feat w
#> 1 sig_a Mlc1 1
#> 2 sig_a Gprc5b 1
#> 3 sig_a Gfap 1
#> 4 sig_b Ednrb 2
#> 5 sig_b Sox9 3
#> 6 sig_b Aqp4 4
g <- createMetafeats(
gobject = g,
feat_clusters = clust_to_use4,
name = "weighted_metagene"
)
#> calculating metafeature: sig_a
#> calculating metafeature: sig_b
# 3. Other statistic transforms can be calculated
# The default and what has been used so far is just finding the mean
g <- createMetafeats(
gobject = g,
feat_clusters = clust_to_use4,
stat = "sum",
expression_values = "raw", # find summed raw expr for these feats
name = "raw_sums"
)
#> calculating metafeature: sig_a
#> calculating metafeature: sig_b
# 4. A custom stat function can also be supplied
# These custom functions must be summary functions, as in, they must
# produce only a single numeric output from many
custom_stat <- function(x) {
if (max(x) == 0) {
return(0)
}
return(mean(x / max(x)))
}
g <- createMetafeats(
gobject = g,
feat_clusters = clust_to_use4,
stat = custom_stat,
expression_values = "raw",
name = "raw_custom"
)
#> calculating metafeature: sig_a
#> calculating metafeature: sig_b
# 5. A rescale of the values to a specified numeric range can be applied
# before the stat score calculation.
g <- createMetafeats(
gobject = g,
feat_clusters = clust_to_use4,
stat = "mean",
expression_values = "normalized",
rescale_to = c(0, 1),
name = "norm_scaled_mean_metafeat"
)
#> calculating metafeature: sig_a
#> calculating metafeature: sig_b
showGiottoSpatEnrichments(g)
#> Spatial unit: aggregate
#>
#> --> Feature type: rna
#>
#> ----> Name cluster_metagene :
#> 1 2 3 4 5
#> <num> <num> <num> <num> <num>
#> 1: 1.014837 0.0000000 0.3160792 0.0000000 0
#> 2: 3.207415 0.9579716 0.6728505 0.2132677 0
#> 3: 3.953661 0.4604975 0.0000000 0.2302488 0
#> 4: 4.093622 0.5886051 0.1962017 0.5886051 0
#> cell_ID
#> <char>
#> 1: 240649020551054330404932383065726870513
#> 2: 274176126496863898679934791272921588227
#> 3: 323754550002953984063006506310071917306
#> 4: 87260224659312905497866017323180367450
#>
#> ----> Name metagene_1 :
#> 1 2 cell_ID
#> <num> <num> <char>
#> 1: 3.493789 3.160792 240649020551054330404932383065726870513
#> 2: 2.132677 2.795048 274176126496863898679934791272921588227
#> 3: 2.633814 7.570115 323754550002953984063006506310071917306
#> 4: 4.446903 3.924034 87260224659312905497866017323180367450
#>
#> ----> Name metagene_2 :
#> sig_a sig_b cell_ID
#> <num> <num> <char>
#> 1: 3.493789 3.160792 240649020551054330404932383065726870513
#> 2: 2.132677 2.795048 274176126496863898679934791272921588227
#> 3: 2.633814 7.570115 323754550002953984063006506310071917306
#> 4: 4.446903 3.924034 87260224659312905497866017323180367450
#>
#> ----> Name metagene_3 :
#> sig_a sig_b cell_ID
#> <num> <num> <char>
#> 1: 3.493789 3.160792 240649020551054330404932383065726870513
#> 2: 2.132677 2.795048 274176126496863898679934791272921588227
#> 3: 2.633814 7.570115 323754550002953984063006506310071917306
#> 4: 4.446903 3.924034 87260224659312905497866017323180367450
#>
#> ----> Name weighted_metagene :
#> sig_a sig_b cell_ID
#> <num> <num> <char>
#> 1: 3.493789 6.321584 240649020551054330404932383065726870513
#> 2: 2.132677 5.590096 274176126496863898679934791272921588227
#> 3: 2.633814 23.041672 323754550002953984063006506310071917306
#> 4: 4.446903 11.772102 87260224659312905497866017323180367450
#>
#> ----> Name raw_sums :
#> sig_a sig_b cell_ID
#> <num> <num> <char>
#> 1: 2 2 240649020551054330404932383065726870513
#> 2: 1 8 274176126496863898679934791272921588227
#> 3: 2 16 323754550002953984063006506310071917306
#> 4: 4 6 87260224659312905497866017323180367450
#>
#> ----> Name raw_custom :
#> sig_a sig_b cell_ID
#> <num> <num> <char>
#> 1: 0.3333333 0.3333333 240649020551054330404932383065726870513
#> 2: 0.3333333 0.3333333 274176126496863898679934791272921588227
#> 3: 0.3333333 0.6666667 323754550002953984063006506310071917306
#> 4: 0.4444444 0.5000000 87260224659312905497866017323180367450
#>
#> ----> Name norm_scaled_mean_metafeat :
#> sig_a sig_b cell_ID
#> <num> <num> <char>
#> 1: 0.3009266 0.6666667 240649020551054330404932383065726870513
#> 2: 0.2139690 0.5895248 274176126496863898679934791272921588227
#> 3: 0.2268554 2.3591938 323754550002953984063006506310071917306
#> 4: 0.4182979 1.2189606 87260224659312905497866017323180367450