ovcCrijns {genefu} | R Documentation |
This function computes subtype scores and risk classifications from gene expression values using teh weights published by Crijns et al.
ovcCrijns(data, annot, hgs, gmap = c("entrezgene", "ensembl_gene_id", "hgnc_symbol", "unigene"), do.mapping = FALSE, verbose = FALSE)
data |
Matrix of gene expressions with samples in rows and probes in columns, dimnames being properly defined. |
annot |
Matrix of annotations with one column named as gmap, dimnames being properly defined. |
hgs |
vector of booleans with TRUE represents the ovarian cancer patients who have a high grade, late stage, serous tumor, FALSE otherwise. This is particularly important for properly rescaling the data. If |
gmap |
character string containing the |
do.mapping |
|
verbose |
|
Note that the original algorithm has not been implemented as it necessitates refitting of the model weights in each new dataset. However the current implementation should give similar results.
score |
Continuous signature scores |
risk |
Binary risk classification, |
mapping |
Mapping used if necessary. |
probe |
If mapping is performed, this matrix contains the correspondence between the gene list (aka signature) and gene expression data. |
Benjamin Haibe-Kains
Crijns APG, Fehrmann RSN, de Jong S, Gerbens F, Meersma G J, Klip HG, Hollema H, Hofstra RMW, te Meerman GJ, de Vries EGE, van der Zee AGJ (2009) "Survival-Related Profile, Pathways, and Transcription Factors in Ovarian Cancer" PLoS Medicine, 6(2):e1000024.
## load the ovsCrijns signature data(sigOvcCrijns) ## load NKI dataset data(nkis) colnames(annot.nkis)[is.element(colnames(annot.nkis), "EntrezGene.ID")] <- "entrezgene" ## compute relapse score ovcCrijns.nkis <- ovcCrijns(data=data.nkis, annot=annot.nkis, gmap="entrezgene", do.mapping=TRUE) table(ovcCrijns.nkis$risk)