multisim {Onassis} | R Documentation |
multisim
This method computes the semantic similarity between samples annotated with different ontology terms from different ontologies
multisim(similarities, annotations, sample1, sample2, aggregating_function) ## S4 method for signature 'list,list,character,character' multisim(similarities, annotations, sample1, sample2, aggregating_function = "mean")
similarities |
a list of Similarity instances, one for each ontology used to annotate the data |
annotations |
a list of annotated data frames obtained using annotateDF or findEntities, one for each ontology |
sample1 |
the name of a sample in annotations |
sample2 |
the name of a sample in annotations |
aggregating_function |
A function used to aggregate the single similarities obtained from each ontology annotation. The function should be applied to a numeric vector. The default value is 'mean' |
The aggregate semantic similarity between the samples sample1 and sample2
ef <- new('EntityFinder') opts <- CMoptions() cell_dict_file <- system.file('extdata', 'sample.cs.obo', package='OnassisJavaLibs') sample_dict <- CMdictionary(cell_dict_file, outputDir=getwd(), synonymType='ALL') samples <- findEntities(ef, system.file('extdata', 'test_samples', 'test_samples.txt', package='Onassis'), outDir=getwd(), multipleDocs=TRUE, configOpt=opts, cmDict=sample_dict) d_dict_file <- system.file('extdata', 'sample.do.obo', package='OnassisJavaLibs') disease_dict <- CMdictionary(d_dict_file, outputDir=getwd(), synonymType='ALL') disease <- findEntities(ef, system.file('extdata', 'test_samples', 'test_samples.txt', package='Onassis'), multipleDocs=TRUE, outDir=getwd(), configOpt=opts, cmDict=disease_dict) cell_sim <- new('Similarity') ontology(cell_sim) <- cell_dict_file disease_sim <- new('Similarity') ontology(disease_sim) <- d_dict_file pairwiseConfig(cell_sim) <- listSimilarities()$pairwiseMeasures[9] pairwiseConfig(disease_sim) <- listSimilarities()$pairwiseMeasures[9] groupConfig(cell_sim) <- listSimilarities()$groupwiseMeasures[3] groupConfig(disease_sim) <- listSimilarities()$groupwiseMeasures[3] similarity <- multisim(list(cell_sim, disease_sim), list(samples, disease), as.character(as.vector(samples[1,1])), as.character(as.vector(samples[5,1])), 'mean')