NMD predictions at transcript and exon levels

Nonsense-mediated decay is an RNA surveillance mechanism which clears transcripts harboring a premature stop codon. These transcripts can be products of alternative splicing events (AS-NMD) which ultimately serve as a post-transcriptional gene regulation mechanism.

The functions below annotates all transcripts in the factRObject for its sensitivity to NMD, based on the canonical exon-junction model. Protein-coding transcripts containing a premature stop codon >50bp upstream of the last exon junction will be annotated as NMD-sensitive.

Sequentially, the `testASNMDevents` function will pinpoint the alternative splicing events that lead to NMD. AS-NMD events are categorised as "stimulating" or "repressing". "Stimulating" events trigger NMD upon its splicing while "repressing" events trigger NMD upon exon skipping.

# S4 method for factR
predictNMD(object, NMD_threshold = 50, verbose = FALSE)

# S4 method for factR
testASNMDevents(object, verbose = FALSE)

Arguments

object

factRObject

NMD_threshold

Minimum distance between PTC and downstream exon-exon junction to trigger NMD (Default: 50)

verbose

Whether to print out messages (Default: FALSE)

Value

factRObject with updated metadata

For `predictNMD`, 4 additional variables are added to the transcript metadata:

  • nmd: "yes" or "no" value as to whether the transcript is NMD-sensitive

  • stop_to_lastEJ: Integer of the distance between the first stop codon to the last exon-junction. Positive values indicate that the stop codon is upstream of the EJ while negative values indicate that the stop codon is downstrea of the EJ

  • num_of_downEJs: Number of EJs downstream of the first stop codon

  • 3'UTR_length: Length of the 3'UTR

For `testASNMDevents`, 2 additional variables are added to the AS metadata:

  • ASNMDtype: Type of AS-NMD event. Can be "Repressing" if skipping of the exon leads to NMD or "Stimulating" if splicing of the exon leads to NMD

  • ASNMD.in.cds: Whether or not the event is found within the CDS or UTR

See also

runfactR

Examples

## Load sample factR object and predict CDS segments
data(factRsample)
factRsample <- buildCDS(factRsample)

## Predict transcript-level NMD sensitivities
factRsample <- predictNMD(factRsample)

## Annotate NMD-causing splicing events
factRsample <- testASNMDevents(factRsample)