Package 'smartid' reference manual

Title:	Scoring and Marker Selection Method Based on Modified TF-IDF
Description:	This package enables automated selection of group specific signature, especially for rare population. The package is developed for generating specifc lists of signature genes based on Term Frequency-Inverse Document Frequency (TF-IDF) modified methods. It can also be used as a new gene-set scoring method or data transformation method. Multiple visualization functions are implemented in this package.
Authors:	Jinjin Chen [aut, cre] (ORCID: <https://orcid.org/0000-0001-7923-5723>)
Maintainer:	Jinjin Chen <[email protected]>
License:	MIT + file LICENSE
Version:	1.9.0
Built:	2026-05-30 09:48:07 UTC
Source:	https://github.com/bioc/smartid

calculate combined score

Description

compute TF (term/feature frequency), IDF (inverse document/cell frequency), IAE (inverse average expression of features) and combine the the final score

Usage

cal_score(
  data,
  tf = c("logtf", "tf"),
  idf = "prob",
  iae = "prob",
  slot = "counts",
  new.slot = "score",
  par.idf = NULL,
  par.iae = NULL,
  return.intermediate = FALSE
)

## S4 method for signature 'AnyMatrix'
cal_score(
  data,
  tf = c("logtf", "tf"),
  idf = "prob",
  iae = "prob",
  par.idf = NULL,
  par.iae = NULL,
  return.intermediate = FALSE
)

## S4 method for signature 'SummarizedExperiment'
cal_score(
  data,
  tf = c("logtf", "tf"),
  idf = "prob",
  iae = "prob",
  slot = "counts",
  new.slot = "score",
  par.idf = NULL,
  par.iae = NULL,
  return.intermediate = FALSE
)
cal_score(
  data,
  tf = c("logtf", "tf"),
  idf = "prob",
  iae = "prob",
  slot = "counts",
  new.slot = "score",
  par.idf = NULL,
  par.iae = NULL,
  return.intermediate = FALSE
)

## S4 method for signature 'AnyMatrix'
cal_score(
  data,
  tf = c("logtf", "tf"),
  idf = "prob",
  iae = "prob",
  par.idf = NULL,
  par.iae = NULL,
  return.intermediate = FALSE
)

## S4 method for signature 'SummarizedExperiment'
cal_score(
  data,
  tf = c("logtf", "tf"),
  idf = "prob",
  iae = "prob",
  slot = "counts",
  new.slot = "score",
  par.idf = NULL,
  par.iae = NULL,
  return.intermediate = FALSE
)

Arguments

data

an expression object, can be matrix or SummarizedExperiment

tf

a character, specify the TF method to use, can be "tf" or "logtf"

idf

a character, specify the IDF method to use. Available methods can be accessed using idf_iae_methods()

iae

a character, specify the IAE method to use. Available methods can be accessed using idf_iae_methods()

slot

a character, specify which slot to use when data is se object, optional, default 'counts'

new.slot

a character, specify the name of slot to save score in se object, optional, default 'score'

par.idf

other parameters for specified IDF methods

par.iae

other parameters for specified IAE methods

return.intermediate

logical, if TRUE also return or store the intermediate tf, idf and iae matrices. Defaults to FALSE since these objects have the same dimension as the input expression matrix and can dominate memory usage on large datasets. Set to TRUE to restore the pre-1.7.3 behavior where intermediates were kept in metadata() of the SummarizedExperiment output.

Value

A list of matrices or se object containing combined score

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
cal_score(
  data,
  par.idf = list(label = sample(c("A", "B"), 10, replace = TRUE)),
  par.iae = list(label = sample(c("A", "B"), 10, replace = TRUE))
)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
cal_score(
  data,
  par.idf = list(label = sample(c("A", "B"), 10, replace = TRUE)),
  par.iae = list(label = sample(c("A", "B"), 10, replace = TRUE))
)

Calculate score for each feature in each cell

Description

Calculate score for each feature in each cell

Usage

cal_score_init(
  expr,
  tf = c("logtf", "tf"),
  idf = "prob",
  iae = "prob",
  par.idf = NULL,
  par.iae = NULL,
  return.intermediate = FALSE
)
cal_score_init(
  expr,
  tf = c("logtf", "tf"),
  idf = "prob",
  iae = "prob",
  par.idf = NULL,
  par.iae = NULL,
  return.intermediate = FALSE
)

Arguments

expr

a count matrix, features in row and cells in column

tf

a character, specify the TF method to use, can be "tf" or "logtf"

idf

a character, specify the IDF method to use. Available methods can be accessed using idf_iae_methods()

iae

a character, specify the IAE method to use. Available methods can be accessed using idf_iae_methods()

par.idf

other parameters for specified IDF methods

par.iae

other parameters for specified IAE methods

return.intermediate

logical, if TRUE the returned list also contains the intermediate tf, idf and iae objects. Default FALSE keeps only the combined score to avoid the memory overhead of three extra feature-by-cell matrices on large inputs.

Value

a list always containing score; when return.intermediate = TRUE the list additionally contains tf, idf and iae.

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
label <- sample(c("A", "B"), 10, replace = TRUE)
smartid:::cal_score_init(data,
  par.idf = list(label = label),
  par.iae = list(label = label)
)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
label <- sample(c("A", "B"), 10, replace = TRUE)
smartid:::cal_score_init(data,
  par.idf = list(label = label),
  par.iae = list(label = label)
)

compute overall score based on the given marker list

Description

compute overall score based on the given marker list

Usage

gs_score(data, features = NULL, slot = "score", suffix = "score")

## S4 method for signature 'AnyMatrix,ANY'
gs_score(data, features = NULL)

## S4 method for signature 'AnyMatrix,list'
gs_score(data, features = NULL, suffix = "score")

## S4 method for signature 'SummarizedExperiment,ANY'
gs_score(data, features = NULL, slot = "score", suffix = "score")
gs_score(data, features = NULL, slot = "score", suffix = "score")

## S4 method for signature 'AnyMatrix,ANY'
gs_score(data, features = NULL)

## S4 method for signature 'AnyMatrix,list'
gs_score(data, features = NULL, suffix = "score")

## S4 method for signature 'SummarizedExperiment,ANY'
gs_score(data, features = NULL, slot = "score", suffix = "score")

Arguments

data

an expression object, can be matrix or SummarizedExperiment

features

vector or named list, feature names to compute score

slot

a character, specify which slot to use when data is se object, optional, default 'score'

suffix

a character, specify the name suffix to save score when features is a named list

Value

A vector of overall score for each sample

Examples

data <- matrix(rnorm(100), 10, dimnames = list(seq_len(10)))
gs_score(data, features = seq_len(3))
data <- matrix(rnorm(100), 10, dimnames = list(seq_len(10)))
gs_score(data, features = seq_len(3))

Calculate scores of each cell on given features

Description

Calculate scores of each cell on given features

Usage

gs_score_init(score, features = NULL)
gs_score_init(score, features = NULL)

Arguments

score

matrix, features in row and samples in column

features

vector, feature names to compute score

Value

a vector of score

Examples

data <- matrix(rnorm(100), 10, dimnames = list(1:10))
gs_score_init(data, 1:5)
data <- matrix(rnorm(100), 10, dimnames = list(1:10))
gs_score_init(data, 1:5)

standard inverse average expression

Description

standard inverse average expression

Usage

iae(expr, features = NULL, thres = 0)
iae(expr, features = NULL, thres = 0)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

thres

numeric, cell only counts when expr > threshold, default 0

Details

$\mathbf{IAE_i} = log(1+\frac{n}{\hat N_{i,j}+1})$

where $n$ is the total number of cells, $N_{i,j}$ is the counts of feature $i$ in cell $j$ .

Value

a vector of inverse average expression score for each feature

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae(data)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae(data)

inverse average expression using hdbscan cluster as label

Description

inverse average expression using hdbscan cluster as label

Usage

iae_hdb(expr, features = NULL, multi = TRUE, thres = 0, minPts = 2, ...)
iae_hdb(expr, features = NULL, multi = TRUE, thres = 0, minPts = 2, ...)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

multi

logical, if to compute based on binary (FALSE) or multi-class (TRUE)

thres

numeric, cell only counts when expr > threshold, default 0

minPts

integer, minimum size of clusters, default 2. Details in dbscan::hdbscan().

...

parameters for dbscan::hdbscan()

Details

Details as iae_prob().

Value

a matrix of inverse average expression score

Examples

set.seed(123)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae_hdb(data)
set.seed(123)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae_hdb(data)

labeled inverse average expression: IGM

Description

labeled inverse average expression: IGM

Usage

iae_igm(expr, features = NULL, label, lambda = 7, thres = 0)
iae_igm(expr, features = NULL, label, lambda = 7, thres = 0)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

label

vector, group label of each cell

lambda

numeric, hyperparameter for IGM

thres

numeric, cell only counts when expr > threshold, default 0

Details

$\mathbf{IGM_i} = log(1+\lambda\frac{max(mean(N_{i,j\in D})_{k})}{\sum_{k}^{K}(mean(N_{i,j\in D})_{k}*r_{k})+e^{-8}})$

where $\lambda$ is the hyper parameter, $N_{i,j\in D}$ is the counts of feature $i$ in cell $j$ within class $D$ , and $r_k$ is the rank of $mean(N_{i,j\in D})$ .

Value

a vector of inverse gravity moment score for each feature

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae_igm(data, label = sample(c("A", "B"), 10, replace = TRUE))
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae_igm(data, label = sample(c("A", "B"), 10, replace = TRUE))

inverse average expression: max

Description

inverse average expression: max

Usage

iae_m(expr, features = NULL, thres = 0)
iae_m(expr, features = NULL, thres = 0)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

thres

numeric, cell only counts when expr > threshold, default 0

Details

$\mathbf{IAE_{i,j}} = log(1+\frac{max_{\{i^{'}\in j\}}(n_{i^{'}})}{\sum_{j = 1}^{n} max(0, N_{i,j} - threshold)+1})$

where $i$ is the feature $i$ and $i^{'}$ is the feature except $i$ , $N_{i,j}$ is the counts of feature $i$ in cell $j$ , and $n_{i^{'}}$ is $\sum_{j = 1}^{n} sign(N_{i,j} > threshold)$ .

Value

a matrix of inverse average expression score for each feature

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae_m(data)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae_m(data)

labeled inverse average expression: probability based

Description

labeled inverse average expression: probability based

Usage

iae_prob(expr, features = NULL, label, multi = TRUE, thres = 0)
iae_prob(expr, features = NULL, label, multi = TRUE, thres = 0)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

label

vector, group label of each cell

multi

logical, if to compute based on binary (FALSE) or multi-class (TRUE)

thres

numeric, cell only counts when expr > threshold, default 0

Details

$\mathbf{IAE_{i,j}} = log(1+\frac{mean(N_{i,j\in D})}{max(mean(N_{i,j\in \hat D}))+ e^{-8}}*mean(N_{i,j\in D}))$

where $N_{i,j\in D}$ is the counts of feature $i$ in cell $j$ within class $D$ , and $\hat D$ is the class except $D$ .

Value

a matrix of inverse average expression score

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae_prob(data, label = sample(c("A", "B"), 10, replace = TRUE))
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae_prob(data, label = sample(c("A", "B"), 10, replace = TRUE))

labeled inverse average expression: relative frequency

Description

labeled inverse average expression: relative frequency

Usage

iae_rf(expr, features = NULL, label, multi = TRUE, thres = 0)
iae_rf(expr, features = NULL, label, multi = TRUE, thres = 0)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

label

vector, group label of each cell

multi

logical, if to compute based on binary (FALSE) or multi-class (TRUE)

thres

numeric, cell only counts when expr > threshold, default 0

Details

$\mathbf{IAE} = log(1+\frac{mean(N_{i,j\in D})}{max(mean(N_{i,j\in \hat D}))+ e^{-8}})$

where $N_{i,j\in D}$ is the counts of feature $i$ in cell $j$ within class $D$ , and $\hat D$ is the class except $D$ .

Value

a matrix of inverse average expression score

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae_rf(data, label = sample(c("A", "B"), 10, replace = TRUE))
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae_rf(data, label = sample(c("A", "B"), 10, replace = TRUE))

inverse average expression using standard deviation (SD)

Description

inverse average expression using standard deviation (SD)

Usage

iae_sd(expr, features = NULL, log = FALSE, thres = 0)
iae_sd(expr, features = NULL, log = FALSE, thres = 0)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

log

logical, if to do log-transformation

thres

numeric, cell only counts when expr > threshold, default 0

Details

$\mathbf{IAE} = log(1+sd(tf_{i})*\frac{n}{\sum_{j=1}^{n}max(0,N_{i,j})+1})$

where $tf_i$ is the term frequency of feature $i$ , see details in tf(), $n$ is the total number of cells and $N_{i,j}$ is the counts of feature $i$ in cell $j$ .

Value

a vector of inverse average expression score for each feature

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae_sd(data)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::iae_sd(data)

standard inverse cell frequency

Description

standard inverse cell frequency

Usage

idf(expr, features = NULL, thres = 0)
idf(expr, features = NULL, thres = 0)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

thres

numeric, cell only counts when expr > threshold, default 0

Details

$\mathbf{IDF_i} = log(1+\frac{n}{n_i+1})$

where $n$ is the total number of cells, $n_i$ is the number of cells containing feature i.

Value

a vector of inverse cell frequency score for each feature

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf(data)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf(data)

inverse document frequency using hdbscan cluster as label

Description

inverse document frequency using hdbscan cluster as label

Usage

idf_hdb(expr, features = NULL, multi = TRUE, thres = 0, minPts = 2, ...)
idf_hdb(expr, features = NULL, multi = TRUE, thres = 0, minPts = 2, ...)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

multi

logical, if to compute based on binary (FALSE) or multi-class (TRUE)

thres

numeric, cell only counts when expr > threshold, default 0

minPts

integer, minimum size of clusters, default 2. Details in dbscan::hdbscan().

...

parameters for dbscan::hdbscan()

Details

Details as idf_prob().

Value

a matrix of inverse cell frequency score

Examples

set.seed(123)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf_hdb(data)
set.seed(123)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf_hdb(data)

Get names of available IDF and IAE methods

Description

Returns a named vector of IDF/IAE methods

Usage

idf_iae_methods()
idf_iae_methods()

Value

names of methods implemented

Examples

idf_iae_methods()
idf_iae_methods()

labeled inverse cell frequency: IGM

Description

labeled inverse cell frequency: IGM

Usage

idf_igm(expr, features = NULL, label, lambda = 7, thres = 0)
idf_igm(expr, features = NULL, label, lambda = 7, thres = 0)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

label

vector, group label of each cell

lambda

numeric, hyperparameter for IGM

thres

numeric, cell only counts when expr > threshold, default 0

Details

$\mathbf{IGM_i} = log(1+\lambda\frac{max(n_{i,j\in D})_{k}}{\sum_{k}^{K}((n_{i,j\in D})_{k}*r_{k})+e^{-8}})$

where $\lambda$ is the hyper parameter, $n_{i,j\in D}$ is the number of cells containing feature $i$ in class $D$ , $r_k$ is the rank of $n_{i,j\in D}$ .

Value

a vector of inverse gravity moment score for each feature

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf_igm(data, label = sample(c("A", "B"), 10, replace = TRUE))
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf_igm(data, label = sample(c("A", "B"), 10, replace = TRUE))

inverse cell frequency: max

Description

inverse cell frequency: max

Usage

idf_m(expr, features = NULL, thres = 0)
idf_m(expr, features = NULL, thres = 0)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

thres

numeric, cell only counts when expr > threshold, default 0

Details

$\mathbf{IDF_{i,j}} = log(\frac{max_{\{i^{'}\in j\}}(n_{i^{'}})}{n_i+1})$

where $i$ is the feature $i$ and $i^{'}$ is the feature except $i$ , $n_i$ is the number of cells containing feature i, and $n_{i^{'}}$ is the number of cells containing feature $i^{'}$ .

Value

a matrix of inverse cell frequency score for each feature

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf_m(data)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf_m(data)

labeled inverse cell frequency: probability based

Description

labeled inverse cell frequency: probability based

Usage

idf_prob(expr, features = NULL, label, multi = TRUE, thres = 0)
idf_prob(expr, features = NULL, label, multi = TRUE, thres = 0)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

label

vector, group label of each cell

multi

logical, if to compute based on binary (FALSE) or multi-class (TRUE)

thres

numeric, cell only counts when expr > threshold, default 0

Details

$\mathbf{IDF_{i,j}} = log(1+\frac{\frac{n_{i,j\in D}}{n_{j\in D}}}{max(\frac{n_{i,j\in \hat D}}{n_{j\in \hat D}})+ e^{-8}}\frac{n_{i,j\in D}}{n_{j\in D}})$

where $n_{i,j\in D}$ is the number of cells containing feature $i$ in class $D$ , $n_{j\in D}$ is the total number of cells in class $D$ , $\hat D$ is the class except $D$ .

Value

a matrix of inverse cell frequency score

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf_prob(data, label = sample(c("A", "B"), 10, replace = TRUE))
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf_prob(data, label = sample(c("A", "B"), 10, replace = TRUE))

labeled inverse cell frequency: relative frequency

Description

labeled inverse cell frequency: relative frequency

Usage

idf_rf(expr, features = NULL, label, multi = TRUE, thres = 0)
idf_rf(expr, features = NULL, label, multi = TRUE, thres = 0)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

label

vector, group label of each cell

multi

logical, if to compute based on binary (FALSE) or multi-class (TRUE)

thres

numeric, cell only counts when expr > threshold, default 0

Details

$\mathbf{IDF_{i,j}} = log(1+\frac{\frac{n_{i,j\in D}}{n_{j\in D}}}{max(\frac{n_{i,j\in \hat D}}{n_{j\in \hat D}})+ e^{-8}})$

where $n_{i,j\in D}$ is the number of cells containing feature $i$ in class $D$ , $n_{j\in D}$ is the total number of cells in class $D$ , $\hat D$ is the class except $D$ .

Value

a matrix of inverse cell frequency score

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf_rf(data, label = sample(c("A", "B"), 10, replace = TRUE))
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf_rf(data, label = sample(c("A", "B"), 10, replace = TRUE))

inverse cell frequency using standard deviation (SD)

Description

inverse cell frequency using standard deviation (SD)

Usage

idf_sd(expr, features = NULL, log = FALSE, thres = 0)
idf_sd(expr, features = NULL, log = FALSE, thres = 0)

Arguments

expr

a matrix, features in row and cells in column

features

vector, feature names or indexes to compute

log

logical, if to do log-transformation

thres

numeric, cell only counts when expr > threshold, default 0

Details

$\mathbf{IDF_i} = log(1+sd(tf_{i})*\frac{n}{n_i+1})$

where $tf_i$ is the term frequency of feature $i$ , see details in tf(), $n$ is the total number of cells and $n_i$ is the number of cells containing feature $i$ .

Value

a vector of inverse cell frequency score for each feature

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf_sd(data)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::idf_sd(data)

select markers using HDBSCAN method

Description

select markers using HDBSCAN method

Usage

markers_hdbscan(
  top_markers,
  column = ".dot",
  s_thres = NULL,
  method = c("max.one", "remove.min"),
  minPts = 5,
  plot = FALSE,
  ...
)
markers_hdbscan(
  top_markers,
  column = ".dot",
  s_thres = NULL,
  method = c("max.one", "remove.min"),
  minPts = 5,
  plot = FALSE,
  ...
)

Arguments

top_markers

output of top_markers()

column

character, specify which column used as group label

s_thres

NULL or numeric, only features with score > threshold will be returned, if NULL will use 2 * average probability as threshold

method

can be "max.one" or "remove.min", if to only keep features in 1st component or return features not in the last component

minPts

integer, minimum size of clusters for dbscan::hdbscan()

plot

logical, if to plot mixture density and hist

...

other params for dbscan::hdbscan()

Value

a list of markers for each group

Examples

data <- matrix(rnorm(100), 10, dimnames = list(1:10))
top_n <- top_markers(data, label = rep(c("A", "B"), 5))
markers_hdbscan(top_n, minPts = 2)
data <- matrix(rnorm(100), 10, dimnames = list(1:10))
top_n <- top_markers(data, label = rep(c("A", "B"), 5))
markers_hdbscan(top_n, minPts = 2)

select markers using mclust EM method

Description

select markers using mclust EM method

Usage

markers_mclust(
  top_markers,
  column = ".dot",
  prob = 0.99,
  s_thres = NULL,
  method = c("max.one", "remove.min"),
  plot = FALSE,
  ...
)
markers_mclust(
  top_markers,
  column = ".dot",
  prob = 0.99,
  s_thres = NULL,
  method = c("max.one", "remove.min"),
  plot = FALSE,
  ...
)

Arguments

top_markers

output of top_markers()

column

character, specify which column used as group label

prob

numeric, probability cutoff for 1st component classification

s_thres

NULL or numeric, only features with score > threshold will be returned, if NULL will use 2 * average probability as threshold

method

can be "max.one" or "remove.min", if to only keep features in 1st component or return features not in the last component

plot

logical, if to plot mixture density and hist

...

other params for mclust::densityMclust()

Value

a list of markers for each group

Examples

data <- matrix(rnorm(100), 10, dimnames = list(1:10))
top_n <- top_markers(data, label = rep(c("A", "B"), 5))
markers_mclust(top_n)
data <- matrix(rnorm(100), 10, dimnames = list(1:10))
top_n <- top_markers(data, label = rep(c("A", "B"), 5))
markers_mclust(top_n)

select markers using mixtools EM method

Description

select markers using mixtools EM method

Usage

markers_mixmdl(
  top_markers,
  column = ".dot",
  prob = 0.99,
  k = 3,
  ratio = 2,
  dist = c("norm", "gamma"),
  maxit = 1e+05,
  plot = FALSE,
  ...
)
markers_mixmdl(
  top_markers,
  column = ".dot",
  prob = 0.99,
  k = 3,
  ratio = 2,
  dist = c("norm", "gamma"),
  maxit = 1e+05,
  plot = FALSE,
  ...
)

Arguments

top_markers

output of top_markers()

column

character, specify which column used as group label

prob

numeric, probability cutoff for 1st component classification

k

integer, number of components of mixtures

ratio

numeric, ratio cutoff of 1st component mu to 2nd component mu, only when ratio > cutoff will return markers for the group

dist

can be one of "norm" and "gamma", specify if to use mixtools::normalmixEM() or mixtools::gammamixEM()

maxit

integer, maximum number of iterations for EM

plot

logical, if to plot mixture density and hist

...

other params for mixtools::normalmixEM() or mixtools::gammamixEM()

Value

a list of markers for each group

Examples

set.seed(1000)
data <- matrix(rnorm(100), 10, dimnames = list(1:10))
top_n <- top_markers(data, label = rep(c("A", "B"), 5))
markers_mixmdl(top_n, k = 3)
set.seed(1000)
data <- matrix(rnorm(100), 10, dimnames = list(1:10))
top_n <- top_markers(data, label = rep(c("A", "B"), 5))
markers_mixmdl(top_n, k = 3)

boxplot of features overall score

Description

boxplot of features overall score

Usage

ova_score_boxplot(data, features, ref.group, label, method = "t.test")
ova_score_boxplot(data, features, ref.group, label, method = "t.test")

Arguments

data

matrix, features in row and samples in column

features

vector, feature names to plot

ref.group

character, reference group name

label

vector, group labels

method

character, statistical test to use, details in ggpubr::stat_compare_means()

Value

ggplot object

Examples

data <- matrix(rnorm(100), 10, dimnames = list(1:10))
ova_score_boxplot(data, 1:5, ref.group = "A", label = rep(c("A", "B"), 5))
data <- matrix(rnorm(100), 10, dimnames = list(1:10))
ova_score_boxplot(data, 1:5, ref.group = "A", label = rep(c("A", "B"), 5))

scale by mean of group mean for imbalanced data

Description

scale by mean of group mean for imbalanced data

Usage

scale_mgm(expr, label, pooled.sd = FALSE)
scale_mgm(expr, label, pooled.sd = FALSE)

Arguments

expr

matrix

label

a vector of group label

pooled.sd

logical, if to use pooled SD for scaling

Details

$z=\frac{x-\frac{\sum_k^{n_D}(\mu_k)}{n_D}}{s}$

where $\mu_k$ is the mean of x in $k^{th}$ class, and $n_D$ is the number of classes, $s$ is the standard deviation of x, when pooled.sd is set to be TRUE, $s$ will be replaced with $s_{pooled}$ , $s_{pooled}=\sqrt{\frac{\sum_k^{n_D}{(n_k-1){s_k}^2}}{\sum_k^{n_D}{n_k}-k}}$

Value

scaled matrix

Examples

scale_mgm(matrix(rnorm(100), 10), label = rep(letters[1:2], 5))
scale_mgm(matrix(rnorm(100), 10), label = rep(letters[1:2], 5))

barplot of processed score

Description

barplot of processed score

Usage

score_barplot(top_markers, column = ".dot", f_list, n = 30)
score_barplot(top_markers, column = ".dot", f_list, n = 30)

Arguments

top_markers

output of top_markers()

column

character, specify which column used as group label

f_list

a named list of markers

n

numeric, number of returned top genes for each group

Value

ggplot object

Examples

data <- matrix(rnorm(100), 10, dimnames = list(1:10))
top_n <- top_markers(data, label = rep(c("A", "B"), 5))
score_barplot(top_n)
data <- matrix(rnorm(100), 10, dimnames = list(1:10))
top_n <- top_markers(data, label = rep(c("A", "B"), 5))
score_barplot(top_n)

scRNA-seq test data of 4 groups simulated by `splatter`.

Description

A SingleCellExperiment object containing 4 groups with each group up-regulated DEGs saved in metadata.

Usage

data(sim_sce_test)
data(sim_sce_test)

Format

A SingleCellExperiment object of 100genes * 400 cells.

Value

SingleCellExperiment

Source

splatter::splatSimulate()

boxplot of split single feature score

Description

boxplot of split single feature score

Usage

sin_score_boxplot(data, features = NULL, ref.group, label, method = "t.test")
sin_score_boxplot(data, features = NULL, ref.group, label, method = "t.test")

Arguments

data

matrix, features in row and samples in column

features

vector, feature names to plot

ref.group

character, reference group name

label

vector, group labels

method

character, statistical test to use, details in ggpubr::stat_compare_means()

Value

faceted ggplot object

Examples

data <- matrix(rnorm(100), 10, dimnames = list(1:10))
sin_score_boxplot(data, 1:2, ref.group = "A", label = rep(c("A", "B"), 5))
data <- matrix(rnorm(100), 10, dimnames = list(1:10))
sin_score_boxplot(data, 1:2, ref.group = "A", label = rep(c("A", "B"), 5))

compute term/feature frequency within each cell

Description

compute term/feature frequency within each cell

Usage

tf(expr, log = FALSE)
tf(expr, log = FALSE)

Arguments

expr

a count matrix, features in row and cells in column

log

logical, if to do log-transformation

Details

$\mathbf{TF_{i,j}}=\frac{N_{i,j}}{\sum_j{N_{i,j}}}$

where $N_{i,j}$ is the counts of feature i in cell j.

Value

a matrix of term/gene frequency. For a dgCMatrix input the returned object preserves sparsity (log1p(0) == 0); dense input returns a dense matrix.

Examples

data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::tf(data)
data <- matrix(rpois(100, 2), 10, dimnames = list(1:10))
smartid:::tf(data)

scale score and return top markers

Description

scale and transform score and output top markers for groups

Usage

top_markers(
  data,
  label,
  n = 10,
  use.glm = TRUE,
  batch = NULL,
  scale = TRUE,
  use.mgm = TRUE,
  softmax = TRUE,
  slot = "score",
  ...
)

## S4 method for signature 'AnyMatrix'
top_markers(
  data,
  label,
  n = 10,
  use.glm = TRUE,
  batch = NULL,
  scale = TRUE,
  use.mgm = TRUE,
  softmax = TRUE,
  slot = "score",
  ...
)

## S4 method for signature 'SummarizedExperiment'
top_markers(
  data,
  label,
  n = 10,
  use.glm = TRUE,
  batch = NULL,
  scale = TRUE,
  use.mgm = TRUE,
  softmax = TRUE,
  slot = "score",
  ...
)
top_markers(
  data,
  label,
  n = 10,
  use.glm = TRUE,
  batch = NULL,
  scale = TRUE,
  use.mgm = TRUE,
  softmax = TRUE,
  slot = "score",
  ...
)

## S4 method for signature 'AnyMatrix'
top_markers(
  data,
  label,
  n = 10,
  use.glm = TRUE,
  batch = NULL,
  scale = TRUE,
  use.mgm = TRUE,
  softmax = TRUE,
  slot = "score",
  ...
)

## S4 method for signature 'SummarizedExperiment'
top_markers(
  data,
  label,
  n = 10,
  use.glm = TRUE,
  batch = NULL,
  scale = TRUE,
  use.mgm = TRUE,
  softmax = TRUE,
  slot = "score",
  ...
)

Arguments

data

an expression object, can be matrix or SummarizedExperiment

label

a vector of group label

n

integer, number of returned top genes for each group

use.glm

logical, if to use stats::glm() to compute group mean score, if TRUE, also compute mean score difference as output

batch

a vector of batch labels, default NULL

scale

logical, if to scale data by row

use.mgm

logical, if to scale data using scale_mgm()

softmax

logical, if to apply softmax transformation on output

slot

a character, specify which slot to use when data is se object, optional, default 'score'

...

params for top_markers_abs() or top_markers_glm()

Value

A tibble with top n feature names, group labels and ordered scores

Examples

data <- matrix(rgamma(100, 2), 10, dimnames = list(1:10))
top_markers(data, label = rep(c("A", "B"), 5))
data <- matrix(rgamma(100, 2), 10, dimnames = list(1:10))
top_markers(data, label = rep(c("A", "B"), 5))

calculate group median, MAD or mean score and order genes based on scores

Description

calculate group median, MAD or mean score and order genes based on scores

Usage

top_markers_abs(
  data,
  label,
  n = 10,
  pooled.sd = FALSE,
  method = c("median", "mad", "mean"),
  scale = TRUE,
  use.mgm = TRUE,
  softmax = TRUE,
  tau = 1
)
top_markers_abs(
  data,
  label,
  n = 10,
  pooled.sd = FALSE,
  method = c("median", "mad", "mean"),
  scale = TRUE,
  use.mgm = TRUE,
  softmax = TRUE,
  tau = 1
)

Arguments

data

matrix, features in row and samples in column

label

a vector of group label

n

integer, number of returned top genes for each group

pooled.sd

logical, if to use pooled SD for scaling

method

character, specify metric to compute, can be one of "median", "mad", "mean"

scale

logical, if to scale data by row

use.mgm

logical, if to scale data using scale_mgm()

softmax

logical, if to apply softmax transformation on output

tau

numeric, hyper parameter for softmax

Value

a tibble with feature names, group labels and ordered processed scores

Examples

data <- matrix(rgamma(100, 2), 10, dimnames = list(1:10))
top_markers_abs(data, label = rep(c("A", "B"), 5))
data <- matrix(rgamma(100, 2), 10, dimnames = list(1:10))
top_markers_abs(data, label = rep(c("A", "B"), 5))

calculate group mean score using glm and order genes based on scores difference

Description

calculate group mean score using glm and order genes based on scores difference

Usage

top_markers_glm(
  data,
  label,
  n = 10,
  family = gaussian(),
  batch = NULL,
  scale = TRUE,
  use.mgm = TRUE,
  pooled.sd = FALSE,
  softmax = TRUE,
  tau = 1
)
top_markers_glm(
  data,
  label,
  n = 10,
  family = gaussian(),
  batch = NULL,
  scale = TRUE,
  use.mgm = TRUE,
  pooled.sd = FALSE,
  softmax = TRUE,
  tau = 1
)

Arguments

data

matrix, features in row and samples in column

label

a vector of group label

n

integer, number of returned top genes for each group

family

family for glm, details in stats::glm()

batch

a vector of batch labels, default NULL

scale

logical, if to scale data by row

use.mgm

logical, if to scale data using scale_mgm()

pooled.sd

logical, if to use pooled SD for scaling

softmax

logical, if to apply softmax transformation on output

tau

numeric, hyper parameter for softmax

Details

When family is gaussian() with the identity link (the default) and the design matrix is full-rank, top_markers_glm() computes all per- gene label coefficients in a single closed-form least-squares solve via Matrix::solve(crossprod(X), crossprod(X, t(data))), avoiding the per-gene glm() loop. For any other family, or a rank-deficient design, the function automatically falls back to the legacy apply(data, 1, glm(...)) path, so results are unchanged for users who pass e.g. family = Gamma() or family = poisson().

Value

a tibble with feature names, group labels and ordered processed scores

Examples

data <- matrix(rgamma(100, 2), 10, dimnames = list(1:10))
top_markers_glm(data, label = rep(c("A", "B"), 5))
data <- matrix(rgamma(100, 2), 10, dimnames = list(1:10))
top_markers_glm(data, label = rep(c("A", "B"), 5))

compute group summarized score and order genes based on processed scores

Description

compute group summarized score and order genes based on processed scores

Usage

top_markers_init(
  data,
  label,
  n = 10,
  use.glm = TRUE,
  batch = NULL,
  scale = TRUE,
  use.mgm = TRUE,
  softmax = TRUE,
  ...
)
top_markers_init(
  data,
  label,
  n = 10,
  use.glm = TRUE,
  batch = NULL,
  scale = TRUE,
  use.mgm = TRUE,
  softmax = TRUE,
  ...
)

Arguments

data

matrix, features in row and samples in column

label

a vector of group label

n

integer, number of returned top genes for each group

use.glm

logical, if to use stats::glm() to compute group mean score, if TRUE, also compute mean score difference as output

batch

a vector of batch labels, default NULL

scale

logical, if to scale data by row

use.mgm

logical, if to scale data using scale_mgm()

softmax

logical, if to apply softmax transformation on output

...

params for top_markers_abs() or top_markers_glm()

Value

a tibble with feature names, group labels and ordered processed scores

Examples

data <- matrix(rgamma(100, 2), 10, dimnames = list(1:10))
top_markers_init(data, label = rep(c("A", "B"), 5))
data <- matrix(rgamma(100, 2), 10, dimnames = list(1:10))
top_markers_init(data, label = rep(c("A", "B"), 5))

Package 'smartid'

Help Index

calculate combined score

Description

Usage

Arguments

Value

Examples

Calculate score for each feature in each cell

Description

Usage

Arguments

Value

Examples

compute overall score based on the given marker list

Description

Usage

Arguments

Value

Examples

Calculate scores of each cell on given features

Description

Usage

Arguments

Value

Examples

standard inverse average expression

Description

Usage

Arguments

Details

Value

Examples

inverse average expression using hdbscan cluster as label

Description

Usage

Arguments

Details

Value

Examples

labeled inverse average expression: IGM

Description

Usage

Arguments

Details

Value

Examples

inverse average expression: max

Description

Usage

Arguments

Details

Value

Examples

labeled inverse average expression: probability based

Description

Usage

Arguments

Details

Value

Examples

labeled inverse average expression: relative frequency

Description

Usage

Arguments

Details

Value

Examples

inverse average expression using standard deviation (SD)

Description

Usage

Arguments

Details

Value

Examples

standard inverse cell frequency

Description

Usage

Arguments

Details