This function takes an object of class iCellR and runs anchor alignment. It's a wrapper for Seurat.
run.anchor(
x = NULL,
method = "base.mean.rank",
top.rank = 500,
gene.list = "character",
data.type = "main",
normalization.method = "LogNormalize",
scale.factor = 10000,
margin = 1,
block.size = NULL,
selection.method = "vst",
nfeatures = 2000,
anchor.features = 2000,
scale = TRUE,
sct.clip.range = NULL,
reduction = c("cca", "rpca"),
l2.norm = TRUE,
dims = 1:30,
k.anchor = 5,
k.filter = 200,
k.score = 30,
max.features = 200,
nn.method = "rann",
eps = 0,
k.weight = 100
)An object of class iCellR.
Choose from "base.mean.rank" or "gene.model", default is "base.mean.rank". If gene.model is chosen you need to provide gene.list.
A number taking the top genes ranked by base mean, default = 500.
A charactor vector of genes to be used for PCA. If "clust.method" is set to "gene.model", default = "my_model_genes.txt".
Choose from "main" and "imputed", default = "main"
Choose from "LogNormalize", "CLR" and "RC". LogNormalize: Feature counts for each cell are divided by the total counts for that cell and multiplied by the scale.factor. This is then natural-log transformed using log1p. CLR: Applies a centered log ratio transformation. RC: Relative counts. Feature counts for each cell are divided by the total counts for that cell and multiplied by the scale.factor. No log-transformation is applied. For counts per million (CPM) set <U+2018>scale.factor = 1e6<U+2019>
Sets the scale factor for cell-level normalization.
If performing CLR normalization, normalize across features (1) or cells (2)
How many cells should be run in each chunk, will try to split evenly across threads
Choose from "vst","mean.var.plot (mvp)","dispersion (disp)".
Number of features to select as top variable features; only used when <U+2018>selection.method<U+2019> is set to <U+2018>'dispersion'<U+2019> or <U+2018>'vst'<U+2019>
A numeric value. This will call <U+2018>SelectIntegrationFeatures<U+2019> to select the provided number of features to be used in anchor finding
Whether or not to scale the features provided. Only set to FALSE if you have previously scaled the features you want to use for each object in the object.list
Numeric of length two specifying the min and max values the Pearson residual will be clipped to
cca: Canonical correlation analysis. rpca: Reciprocal PCA
Perform L2 normalization on the CCA cell embeddings after dimensional reduction
Which dimensions to use from the CCA to specify the neighbor search space
How many neighbors (k) to use when picking anchors
How many neighbors (k) to use when filtering anchors
How many neighbors (k) to use when scoring anchors
The maximum number of features to use when specifying the neighborhood search space in the anchor filtering
Method for nearest neighbor finding. Options include: rann, annoy
Error bound on the neighbor finding algorithm (from RANN)
Number of neighbors to consider when weighting
An object of class iCellR.