These method signatures learn gene set signatures optionally augmented with
(semi-)quantitative information for the prediction of sample and cell identities
in SummarizedExperiment objects.
learnSignatures(se, assay.type = "counts", method = c("PositiveProportionDifference", "PPD"), ...)
| se | An object of class inheriting from " |
|---|---|
| assay.type | A string specifying which assay values to use, e.g., |
| method | Learning method. See section "Learning methods". |
| ... | Additional arguments affecting the learning method. |
A Sets object.
Requires prior cluster membership information. This method computes the proportion of samples positive for each feature in each cluster, and subsequently identifies for each cluster the features showing the maximal difference between the detection rate in the cluster of interest and the detection rate in all other clusters.
# Example data ---- library(SummarizedExperiment) nsamples <- 100 u <- matrix(rpois(20000, 2), ncol=nsamples) rownames(u) <- paste0("Gene", sprintf("%03d", seq_len(nrow(u)))) colnames(u) <- paste0("Cell", sprintf("%03d", seq_len(ncol(u)))) se <- SummarizedExperiment(assays=list(counts=u)) # Example usage ---- se1 <- se colData(se1)[, "cluster"] <- factor(sample(head(LETTERS, 3), ncol(se1), replace=TRUE)) gs <- learnSignatures(se1, method="PositiveProportionDifference", cluster.col="cluster") relations(gs)#> Hits object with 15 hits and 2 metadata columns: #> from to | ProportionPositive minDifferenceProportion #> <integer> <integer> | <numeric> <numeric> #> [1] 1 1 | 0.969697 0.169697 #> [2] 2 1 | 0.848485 0.129735 #> [3] 3 1 | 0.969697 0.125947 #> [4] 4 1 | 1.000000 0.114286 #> [5] 5 1 | 0.909091 0.109091 #> ... ... ... . ... ... #> [11] 11 3 | 0.914286 0.164286 #> [12] 12 3 | 0.942857 0.130357 #> [13] 13 3 | 0.914286 0.126407 #> [14] 14 3 | 0.942857 0.124675 #> [15] 15 3 | 1.000000 0.121212 #> ------- #> nLnode: 15 / nRnode: 3