Skip to contents

Perform UMAP on the Adjacency Matrix of a Milo Object

miloUmap.Rd

This function uses uwot::umap to perform UMAP dimensionality reduction on the adjacency matrix of the KNN graph in a Milo object.

Usage

miloUmap(
  milo,
  slot_name = "UMAP_knngraph",
  n_neighbors = 50L,
  metric = "euclidean",
  min_dist = 0.3,
  ...
)

Arguments

milo

the milo object with knn graph that needed to conduct umap on.

slot_name

character, with default 'UMAP_knngraph'.

  • The slot name in reduceDim where the result store

n_neighbors

integer, with default 50L.

  • the size of local neighborhood (in terms of number of neighboring sample points) used for manifold approximation.

  • Here, the goal is to create large enough neighborhoods to capture the local manifold structure to allow for hypersampling.

metric

character, with default 'euclidean'

  • the choice of metric used to measure distance to find nearest neighbors. Default is 'euclidean'.

min_dist

numeric, with default 0.3

  • the minimum distance between points in the low dimensional space

...

other parameters passed to uwot::umap

Value

milo object with umap reduction

Examples

data(sce_vdj)
sce_vdj <- setupVdjPseudobulk(sce_vdj,
    already.productive = FALSE,
    allowed_chain_status = c("Single pair", "Extra pair")
)
#> Checking productivity from productive_abT_VDJ, productive_abT_VJ ...
#> 7279 of cells filtered
#> checking allowed chain status...
#> 12 of cells filtered
#> VDJ data extraction begin:
#> Parameter extract_cols do not provided, automatically geneterate colnames for extraction.
#> Detect whether colData v_call_abT_VDJ, d_call_abT_VDJ, j_call_abT_VDJ, v_call_abT_VJ, j_call_abT_VJ already exist...
#> Extract main TCR from v_call_abT_VDJ, d_call_abT_VDJ, j_call_abT_VDJ, v_call_abT_VJ, j_call_abT_VJ ...
#> Complete.
#> Filtering cells from v_call_abT_VDJ_main, j_call_abT_VDJ_main, v_call_abT_VJ_main, j_call_abT_VJ_main ...
#> 63 of cells filtered
#> 2646 of cells remain.
# Build Milo Object
milo_object <- miloR::Milo(sce_vdj)
milo_object <- miloR::buildGraph(milo_object, k = 50, d = 20, reduced.dim = "X_scvi")
#> Constructing kNN graph with k:50
milo_object <- miloR::makeNhoods(milo_object, reduced_dims = "X_scvi", d = 20)
#> Checking valid object
#> Running refined sampling with reduced_dim

# Construct UMAP on Milo Neighbor Graph
milo_object <- miloUmap(milo_object)