Skip to content

v0.1.0 - stable version with complete docs and examples
Compare
Choose a tag to compare
@sreichl sreichl released this 15 Jan 13:11
· 55 commits to main since this release
v0.1.0
969319a

features

  • enrichment analysis methods
    • region-sets
      • LOLA: Genomic Locus Overlap Enrichment Analysis is run locally.
      • GREAT using rGREAT: Genomic Regions Enrichment of Annotations Tool is queried remotely (requires a working internet connection).
    • gene-sets
      • over-representation analysis (ORA) using GSEApy enrich() function performs Fisher’s exact test (i.e., hypergeometric test) and is run locally.
      • preranked gene-set enrichment analysis (preranked GSEA) using GSEApy prerank() function performs preranked GSEA and is run locally.

Note: All genomic region sets are subjected to gene-set ORA, leveraging region-gene associations of each query, and background region-set obtained using GREAT. Thereby, an extended region-set enrichment perspective can be gained by querying databases, that are not supported by region-based tools.

  • resources (databases) for both gene-based analyses are either downloaded (Enrichr) or copied from local JSON or GMT files.

    • all Enrichr databases can be queried (enrichr_dbs).
    • local JSON database files can be queried (local_json_dbs).
    • local GMT database files (e.g., from MSigDB) can be queried (local_gmt_dbs).

  • group aggregation of results per method and database

    • results of all queries belonging to the same group are aggregated per method and database.
    • a filtered version taking the union of all statistically significant terms per query is also saved.

  • visualization

    • region/gene-set specific enrichment dot plots are generated for each query, method, and database combination where the top terms are ranked (along the y-axis) by the mean rank of statistical significance, effect-size, and overlap with the goal to make the results more balanced and interpretable.
    • group summary/overview
      • the union of the most significant terms per query, method, and database within a group is determined.
      • their effect-size and statistical significance are visualized as hierarchically clustered heatmaps.
      • a hierarchically clustered bubble plot encoding both effect-size and significance is provided.

docuemntation

  • complete documentation of used software, all features, and methods
  • a minimal example to test all supported features
  • external resources
Assets 2
Loading