proActiv release 0.99.0

[jleechung]

Calculating Promoter Activity

• proActiv function
  • Wrapper function proActiv calls the functions to calculate junctions counts and estimate promoter activity
  • Compulsory arguments file paths files and promoter annotation object promoterAnnotation
  • Optional arguments fileLabels and ncores; genome argument is required if input is a BAM file
  • Additional argument condition can be supplied - proActiv will summarize promoter counts and activity across conditions if this is the case
  • proActiv returns a summarizedExperiment object
    • assays (4) - promoterCounts, normalizedPromoterCounts, absolutePromoterActivity, relativePromoterActivity
    • metadata (1) - geneExpression
    • rowData - DataFrame - promoterID, geneID, transcriptID, seqname, strand, promoter position, promoter activity summarized by condition, and promoter category by condition
  • mclapply to bplapply

Creating Promoter Annotations

• preparePromoterAnnotation function

  • Builds promoter annotation
  • Compulsory arguments - species and either txdb or file ; file can be path to a GTF or a TxDb
  • Improved efficiency in generating annotations without parallelization (~2 mins for hg38 annotation)
  • Main improvements for helper functions getIntronRanks and getReducedExons
  • Combine 4 helper functions in annotation-data into one function to avoid repeated calls to certain helper functions which are time-expensive
  • To comply with Bioconductor guidelines on package size, promoter annotation is shrunk
  • From 4 slots to 3 slots - intronRanges, promoterIdMapping, promoterCoordinates
  • Metadata not necessary for running proActiv is removed

• Promoter annotations for multiple genomes exported to user

  • hg19 / hg38
  • mm9 / mm10

Plotting promoter activity

• plotPromoters function visualizes promoter activity across conditions with Gviz
  • Compulsory arguments result (summarizedExperiment object returned by proActiv), gene of interest and annotation object - either txdb or GRangesList of exonsByTranscripts ranges
  • Allows users to visualize promoter activity and identify instances of alternative promoter usage

Vignette

• proActiv now comes with a vignette, documenting a complete step-by-step workflow in identifying active and alternative promoter usage
• Includes sections:
  • Running proActiv
  • Creating promoter annotations
  • Identifying alternative promoter usage with DEXSeq
  • Visualizations of promoter activity:
    • plotPromoters
    • Promoter category proportions across conditions
    • Major-minor promoters by position (5' -> 3')
    • Major promoter activity vs. gene expression
    • t-SNE plot with active promoters

Tests

• Include units tests for:
  • annotation-data.R / annotation-data-helper.R
  • junction-read-count.R
  • estimate-promoter-activity.R
  • proActiv.R

Reexports

• Reexport certain imported functions from other packages loaded with proActiv namespace - primarily for use in vignette (e.g. getters for summarizedExperiment rowData, assays...)

extdata

• Vignette directory
  • Annotations directory
    • Contains genomic annotation data to be used in the vignette. including
      • gencode.v34.annotations subsetted to chromosome 22 (gtf, sqlite)
      • RAP1GAP data for visualization with plotPromoters
  • Junctions directory
    • Contains STAR junction files from SGNEx data (3 x A549, 3 x HepG2)
• Testdata directory
  • data used for testing (incl. subsetted bam files, star and tophat2 junction files, and promoter annotations)