MVN_CV: Imputation and Construction of High-Resolution Spatial Maps...

View source: R/CARD.imputation.R

MVN_CVR Documentation

Imputation and Construction of High-Resolution Spatial Maps for Cell Type Composition and Gene Expression by the spatial correlation structure between original spatial locations and new grided spatial locations

Description

Imputation and Construction of High-Resolution Spatial Maps for Cell Type Composition and Gene Expression by the spatial correlation structure between original spatial locations and new grided spatial locations

Usage

MVN_CV(
  Vtrain,
  locationOrig,
  trainInd,
  testInd,
  B,
  Xinput_norm,
  Optimalb,
  OptimalPhi,
  lambda,
  ineibor
)

Arguments

Vtrain

Matrix, estimated V matrix from CARD

locationOrig

Data frame, spatial location data frame of the original spatial resolved transcriptomics dataset, stored in the CARD_object@spatial_location

trainInd

Vector, index of the original spatial locations

testInd

Vector, index of the newly grided spatial locations

B

Matrix, used in the deconvolution as the reference basis matrix

Xinput_norm

Matrix, used in the deconvolution as the normalized spatial count data

Optimalb

Vector, vector of the intercept for each cel type estimated based on the original spatial resolution

OptimalPhi

Numeric, the optimal phi value stored in CARD_object

lambda

Vector, vector of cell type specific scalar in the CAR model

ineibor

Numeric, number of neighbors used in the imputation on newly grided spatial locations, default is 10.

Value

Return a list with the imputed Cell type composition Vtest matrix on the newly grided spatial locations and predicted normalized gene expression


YingMa1993/CARD documentation built on Feb. 14, 2024, 12:24 p.m.