methods: Get and set methods for Milo objects

In MikeDMorgan/miloR: Differential neighbourhood abundance testing on a graph

Get and set methods for Milo objects

Description

Get and set methods for Milo object slots. Generally speaking these methods are used internally, but they allow the user to assign their own externally computed values - should be used with caution.

Value

See individual methods for return values

Getters

In the following descriptions x is always a Milo object.

graph(x):

Returns an igraph object representation of the KNN-graph, with number of vertices equal to the number of single-cells.

nhoodDistances(x):

Returns a list of sparse matrix of cell-to-cell distances between nearest neighbours, one list entry per neighbourhood. Largely used internally for computing the k-distance weighting in graphSpatialFDR.

nhoodCounts(x):

Returns a NxM sparse matrix of cell counts in each of N neighbourhoods with respect to the M experimental samples defined.

nhoodExpression(x):

Returns a GxN matrix of gene expression values.

nhoodIndex(x):

Returns a list of the single-cells that are the neighbourhood indices.

nhoodReducedDim(x):

Returns an NxP matrix of reduced dimension positions. Either generated by projectNhoodExpression(x) or by providing an NxP matrix (see setter method below).

nhoods(x):

Returns a sparse matrix of CxN mapping of C single-cells toN neighbourhoods.

nhoodGraph(x):

Returns an igraph object representation of the graph of neighbourhoods, with number of vertices equal to the number of neighbourhoods.

nhoodAdjacency(x):

Returns a matrix of N by N neighbourhoods with entries of 1 where neighbourhods share cells, and 0 elsewhere.

Setters

In the following descriptions x is always a Milo object.

graph(x) <- value:

Populates the graph slot with value - this should be a valid graph representation in either igraph or list format.

nhoodDistances(x) <- value:

Replaces the internally comptued neighbourhood distances. This is normally computed internally during graph building, but can be defined externally. Must be a list with one entry per neighbourhood containing the cell-to-cell distances for the cells within that neighbourhood.

nhoodCounts(x) <- value:

Replaces the neighbourhood counts matrix. This is normally computed and assigned by countCells, however, it can also be user-defined.

nhoodExpression(x) <- value:

Replaces the nhoodExpression slot. This is calculated internally by calcNhoodExpression, which calculates the mean expression. An alternative summary function can be used to assign an alternative in this way.

nhoodIndex(x) <- value:

Replaces the list of neighbourhood indices. This is provided purely for completeness, and is usually only set internally in makeNhoods.

nhoodReducedDim(x) <- value:

Replaces the reduced dimensional representation or projection of neighbourhoods. This can be useful for externally computed projections or representations.

nhoods(x) <- value:

Replaces the neighbourhood matrix. Generally use of this function is discouraged, however, it may be useful for users to define their own bespoke neighbourhoods by some means.

nhoodGraph(x) <- value:

Populates the nhoodGraph slot with value - this should be a valid graph representation in either igraph or list format.

nhoodAdjacency(x) <- value:

Populates the nhoodAdjacency slot with value - this should be a N by N matrix with elements denoting which neighbourhoods share cells

Miscellaneous

A collection of non-getter and setter methods that operate on Milo objects.

show(x):

Prints information to the console regarding the Milo object.

Author(s)

Mike Morgan

Examples

example(Milo, echo=FALSE)
show(milo)

MikeDMorgan/miloR documentation built on Feb. 29, 2024, 6:20 p.m.