Diritchlet-multinomial differential analysis of splicing events (DM-DASE). The workflow includes pre-computing splicing events counts with JuncBASE or percent-spliced in (PSI) values with MESA or both for comparison or benchmarking analyses. Program performs differential transcript and gene usage (DU) or differential splicing (DS) analysis at the event- and gene-level in either JuncBASE or MESA or DRIMSeq formatted data tables by computing both gene- and transcript-level DM distributions (where each transcript ratio is modeled separately) using DRIMSeq. Program also includes the options to perform batch correction DU or DS and transcript usage QTL (tuQTL) analysis.
You can find documentation here
- Formats input data to DRIMSeq format
- Filters funtional AS events by mapping them to reference. Events are considered functional if there is a transcript annotation associated with the event region, this allows to select for specific genomic features (optional) and reduces the computational cost of downstream analyses.
- Quantifies proportions for global, known and novel AS events (JuncBASE only).
- Performs either DU or DS global and/or specific comparison analyses
- Performs gene and transcript level p-value adjustment using a two-stage test from stageR
- Calculates ΔPSI and LFC values from fitted transcript proportions from all MT vs WT conditions (for global analysis)
- If performing comparison analysis, model calculates ΔPSI and LFC values from fitted transcript proportions from samples in condition1 vs samples in condition2
- Performs batch correction DU or DS analysis (optional)
- Annotates AS events with differential usage or splicing (both JuncBASE and/or MESA)
- Quantifies significant AS event proportions from DU or DS results (JuncBASE only)
- Perform enrichment/pathway analysis on significant AS events
