NEWS

---

GSEABenchmarkeR 1.8.0

• Extended support for benchmarking user-defined inputs: - mixing of pre-defined and user-defined enrichment methods (functions 'runEA' and 'evalTypeIError') - simplified passing on of additional arguments to user-defined enrichment methods for functions 'runEA' and 'evalTypeIError'
• The TCGA RNA-seq compendium can now also be obtained via curatedTCGAData using 'loadEData("tcga", mode = "cTD")'

GSEABenchmarkeR 1.6.0

• New function 'evalTypeIError': type I error evalution by sample permutation - evaluation of >= 1 enrichment methods on >= 1 expression datasets - support for splitting permutations into blocks of defined size + invoking parallel evaluation of the partitions
• New function 'evalRandomGS' for evaluation of random gene sets: - estimates proportion of rejected null hypotheses (= fraction of significant gene sets) of an enrichment method when applied to random gene sets of defined size - evaluation of >= 1 enrichment methods on an expression dataset of choice
• New argument ‘method' to the 'evalRelevance' function for the evaluation of phenotype relevance of gene set rankings, choices include: - "wsum": computes a weighted sum of the relevance scores (default), - "auc": performs a ROC/AUC analysis based on the ROCR package, - "cor": computes a standard correlation such as Spearman’s rank correlation, - a user-defined function for customized behaviors.
• New function 'metaFC' for summarizing fold changes of individual datasets across a compendium of expression datasets
• New functions 'plotDEDistribution' and 'plotNrSamples' for exploring differential expression and sample size across a compendium of expression datasets
• Extended support for user-defined benchmarking inputs including simplified plug-in of user-defined enrichment methods (thanks to Marcel Ramos @LiNk-NY)