Integrative Pathway Enrichment Analysis of Multivariate Omics
Framework for analysing multiple omics datasets in the context of molecular pathways, biological processes and other types of gene sets. The package uses p-value merging to combine gene- or protein-level signals, followed by ranked hypergeometric tests to determine enriched pathways and processes. This approach allows researchers to interpret a series of omics datasets in the context of known biology and gene function, and discover associations that are only apparent when several datasets are combined. The package is part of the following publication: Integrative Pathway Enrichment Analysis of Multivariate Omics Data. Paczkowska M^, Barenboim J^, Sintupisut N, Fox NS, Zhu H, Abd-Rabbo D, Mee MW, Boutros PC, PCAWG Drivers and Functional Interpretation Working Group; Reimand J, PCAWG Consortium. Nature Communications (2020) <10.1038>.10.1038>
- Add a geneset.filter option to activeDriverPW to filter the GMT by geneset size
- Add vignettes
- Export merge_p_values method
- Change column contribution method. Column contribution is now reported as the
log-fold-change when the column is excluded. ie, -log10(p_val_with_column / p_val_without_column)
- No longer raises an error if no significant terms are found and
is false. Instead returns an empty
data.table and issues a warning.
- Issue a warning if genes are filtered out for not being found in the background
- Change default p-value adjustment method to "holm"
- If return.all==FALSE, calculates columnContribution only for terms that will
returned. Speeds up runtime by roughly a factor of 2
- Add "none" option to p-value adjustment methods
- New implementation of orderedHypergeometric function which is several times faster.
- Fix Brown's method when all p-values in a column are the same
- Fix in orderedHypergeometric which added an extra NA to the
to small errors in the calculated p-value
- Fixed another bug in orderedHypergeometric which added an extra NA to the
complement in some cases, leading to small errors in the calculated p-value