splicelogic: splicing events from transcript sets

splicelogic turns sets of transcripts into discrete splicing events. Unlike event-based tools that work at the junction level, splicelogic operates on whole transcript structures: within each gene it compares two groups of transcripts and all of their exons, so events are derived with full isoform context. It detects skipped exons, included exons, mutually exclusive exons, retained and spliced introns, alternative 5' and 3' splice sites, and alternative transcription start and end sites.

splicelogic allows the two groups of transcripts to be compared to be defined upstream in one of two ways. They can come from an explicit partition of the transcripts into two sets, e.g. a set of novel transcripts against a set of annotated reference transcripts, or transcripts found in disease samples relative to those found in control samples. Or they can come from a differential transcript usage (DTU) analysis, which gives each transcript an effect estimate whose sign defines the comparison. Because it takes transcript-level results tables as input, splicelogic is compatible with any upstream DTU method (including DRIMSeq, DEXSeq, satuRn, and edgeR), supporting flexible experimental designs.

splicelogic operates on exon-level data stored as GRanges objects within R/Bioconductor. Given a set of exons carrying the gene ID, transcript ID and exon rank, splicelogic can be used to identify a variety of splicing events. See the vignette for more details. splicelogic works well within the tidyomics framework. Event results including the exons and introns underlying each event can be further manipulated downstream using plyranges.

How to install

splicelogic is available from Bioconductor:

if (!require("BiocManager", quietly = TRUE))
  install.packages("BiocManager")
BiocManager::install("splicelogic")

The most recent development version can be installed from the devel branch on GitHub:

devtools::install_github("thelovelab/splicelogic")

Quick start

Finding events from transcript sets

# prepare exons from two sets of transcripts
exons <- prepare_exons_by_partition(
  up = <GRanges OF EXONS, OR tx IDs>,
  down = <GRanges OF EXONS, OR tx IDs>
) |>
  preprocess(coef_col = "estimate")

# find skipped exons
skipped <- exons |> find_se()

# find all splicing events
all_events <- exons |> find_all_events()

Finding events from DTU results

# prepare exons from a TxDb and DTU results
exons <- prepare_exons(
  txdb = <A TxDb OBJECT>,
  dtu_table = <DTU_TABLE>,
  coef_col = "estimate"
)

# preprocess for further analysis
exons <- preprocess(exons, coef_col = "estimate")

# find skipped exons
skipped <- exons |> find_se()

# find all splicing events
all_events <- exons |> find_all_events()

Future directions

• Support detection of additional event types, such as consecutive skipped exons.
• Extraction and labelling of the specific splice junctions associated with each event, adding metadata columns such as the donor–acceptor dinucleotide sequence (e.g. AG-GT) and a logical indicating whether the junction is canonical, for downstream interpretation.
• Custom plotting functions to easily visualize the results, showing the exons involved in each event alongside the transcripts they come from and the event annotation.
• Support detection of alternative UTR events (alternative 5' and 3' UTRs), when the reference annotation includes UTR coordinates (e.g. GENCODE), building on the alternative transcription start and end sites reported by find_atss() and find_ates().
• Facilitating RNA-binding protein (RBP) motif detection.
• Facilitating interpretation of downstream structural consequences.

Feedback

We would love to hear your feedback. Please post to an Issue on GitHub.

Funding

splicelogic is supported by NHGRI R01-HG009937, and the Wellcome Trust as part of the EOSS program.