WISP.getWeight: WISP.getWeight

In cit-bioinfo/WISP: Weighted In Silico Pathology

Description

Pure population weight estimations for each sample.

Usage

WISP.getWeight(data, centro, scaling = c("none", "scale", "center")[1], cutoff_gobalFtest = 0.05, Rsquared_cutoff = 0.5, cutoff_ttest_weights = 0.05, sum_LessThanOne = TRUE) 

Arguments

data	data.frame of gene expression with sample in column.
centro	data.frame of centroids given by WISP.getPureClass.
scaling	type of data and centroids scaling: none, centered or scaled.
cutoff_gobalFtest	p-value threshold of the gobal F-test to considere the model as OK or LIMIT for a sample.
Rsquared_cutoff	Rsquared cutoff to considere the model as OK or LIMIT for a sample.
cutoff_ttest_weights	p-value threshold of the t-test for each weight.
sum_LessThanOne	sum of coefficients less that one (TRUE) or equal to one (FALSE).

Value

output

dataframe containing the weight of each pure populations, Distance to model, F.test Pvalue, adjusted R squared and for each weight the Confidence Interval and t.test Pvalue. Last column correpond to a WARNING based on the F-test Pvalue and the adjusted R squared cutoff.

Author(s)

Yuna Blum, Aurelien de Reynies and Sylvie Job

cit-bioinfo/WISP documentation built on June 8, 2019, 8:41 p.m.