ISS_clustCompare: Compare among ISS clusters based on Random Forest algorithm

In mashranga/MolDia: Single cell In-Situ sequencing (ISS) data analysis

Description

Compare two cluster and find their similarity (Supervised).

Usage

ISS_clustCompare(preditorData, prediction, ntree = 100,
  method = "prob")

Arguments

preditorData	Data to be used to compare. Input data in class MolDiaISS. Output of readISS and must be clustered.
prediction	Data that need to compare. Input data in class MolDiaISS. Output of readISS and must be clustered.
ntree	Number of trees to grow. randomForest. See This should not be set to too small a number, to ensure that every input row gets predicted at least a few times.
method	Methods to return comarison results. Possible values are "vote" and "prob". Defaut is "prob".

Details

This function will be apply to compare Two cluster by Random forest algorithm.

Note

This function need optimization to run properly. Work fine with small data but not with big data .

Examples

## Read data: Left and right HC
left_hypo  <- readISS(file = system.file("extdata", "Hypocampus_left.csv", package="MolDia"),
                  cellid = "CellId",centX = "centroid_x", centY = "centroid_y")
right_hypo <- readISS(file = system.file("extdata", "Hypocampus_right.csv", package="MolDia"),

                                     cellid = "CellId",centX = "centroid_x", centY = "centroid_y")
## Data preprocessing
left_hypo  <- ISS_preprocess(data = left_hypo, normalization.method = "LogNormalize",
                               do.scale = TRUE, do.center = TRUE)
right_hypo <- ISS_preprocess(data = right_hypo, normalization.method = "LogNormalize",
                               do.scale = TRUE, do.center = TRUE)

## Cluster data based on SEURAT pipeline
left_hypo   <- ISS_cluster(data =  left_hypo, pc = 0.9, resolution = 0.05)
right_hypo  <- ISS_cluster(data = right_hypo, pc = 0.9, resolution = 0.05)

## Cluster compare
comapre_left_right <- ISS_clustCompare(preditorData = left_hypo, prediction = right_hypo, ntree = 10)
comapre_right_left <- ISS_clustCompare(preditorData = right_hypo, prediction = left_hypo, ntree = 10)

mashranga/MolDia documentation built on May 26, 2019, 9:36 a.m.