Skip to contents

Impute the peak counts from gene expression data referring to an ATAC dataset after integration

Source: R/ATAC.R
imputeKNN.Rd

This function is designed for creating peak data for a dataset with only gene expression. This function uses quantile normalized cell factor loading to find nearest neighbors between cells from the queried dataset (without peak) and cells from reference dataset (with peak). And then impute the peak for the former basing on the weight. Therefore, the reference dataset selected must be of "atac" modality setting.

Usage

imputeKNN(
  object,
  reference,
  queries = NULL,
  nNeighbors = 20,
  weight = TRUE,
  norm = TRUE,
  scale = FALSE,
  verbose = getOption("ligerVerbose", TRUE),
  ...,
  knn_k = nNeighbors
)

Arguments

object

liger object with aligned factor loading computed in advance.

reference

Name of a dataset containing peak data to impute into query dataset(s).

queries

Names of datasets to be augmented by imputation. Should not include reference. Default NULL uses all datasets except the reference.

nNeighbors

The maximum number of nearest neighbors to search. Default 20.

weight

Logical. Whether to use KNN distances as weight matrix. Default FALSE.

norm

Logical. Whether to normalize the imputed data. Default TRUE.

scale

Logical. Whether to scale but not center the imputed data. Default TRUE.

verbose

Logical. Whether to show information of the progress. Default getOption("ligerVerbose") or TRUE if users have not set.

...

Optional arguments to be passed to normalize when norm = TRUE.

knn_k

Deprecated. See Usage section for replacement.

Value

The input object where queried ligerDataset

objects in datasets slot are replaced. These datasets will all be converted to ligerATACDataset class with an additional slot rawPeak to store the imputed peak counts, and normPeak for normalized imputed peak counts if norm = TRUE.

Examples

bmmc <- normalize(bmmc)
#> ℹ Normalizing datasets "rna"
#> ℹ Normalizing datasets "atac"
#> ✔ Normalizing datasets "atac" ... done
#> 
#> ℹ Normalizing datasets "rna"

#> ✔ Normalizing datasets "rna" ... done
#> 
bmmc <- selectGenes(bmmc, datasets.use = "rna")
#> Warning: The `datasets.use` argument of `selectGenes.liger()` is deprecated as of rliger
#> 1.99.0.
#> ℹ Please use the `useDatasets` argument instead.
#> ℹ The deprecated feature was likely used in the rliger package.
#>   Please report the issue at <https://github.com/welch-lab/liger/issues>.
#> ℹ Selecting variable features for dataset "rna"
#> ✔ ... 83 features selected out of 172 shared features.
#> ✔ Finally 83 shared variable features are selected.
bmmc <- scaleNotCenter(bmmc)
#> ℹ Scaling dataset "rna"
#> ✔ Scaling dataset "rna" ... done
#> 
#> ℹ Scaling dataset "atac"
#> ✔ Scaling dataset "atac" ... done
#> 
if (requireNamespace("RcppPlanc", quietly = TRUE)) {
    bmmc <- runINMF(bmmc, k = 20)
    bmmc <- quantileNorm(bmmc)
    bmmc <- normalizePeak(bmmc)
    bmmc <- imputeKNN(bmmc, reference = "atac", queries = "rna")
}
#> ℹ Using largest dataset of recommended type as reference: "rna" with 340 cells
#> ℹ Normalizing peak of dataset: "atac"
#> ✔ Normalizing peak of dataset: "atac" ... done
#> 
#> ℹ Imputing 1 query dataset: "rna"
#> ℹ from reference dataset: "atac"
#> ℹ Normalizing peak of dataset: "rna"
#> ✔ Normalizing peak of dataset: "rna" ... done
#>