The Junction Usage Model (JUM)

JUM is a tool that offers a completely annotation-free approach to analyze alternative pre-mRNA splicing patterns that are specific to a tissue or sample of interest from RNA-seq datasets.
- Stable performance package for V3.0.0: https://github.com/qqwang-berkeley/JUM/releases/tag/v3.0.0
- Complete manual for V3.0.0 and V2.0.2: https://github.com/qqwang-berkeley/JUM/wiki/A-Hitchhiker's-guide-to-JUM-(version-2.0.2--and-3.0.0)
- Legacy release for V2.0.2: https://github.com/qqwang-berkeley/JUM/releases/tag/JUM_v2.0.2
- For differential AS analysis among multiple experimental conditions see: https://github.com/qqwang-berkeley/JUM/wiki/6.-Running-JUM-for-differential-AS-analysis-among-multiple-experimental-conditions-(multiple-tissues,-patients,-time-series-experiments-etc.)
JUM reports alternative splicing events in six categories: cassette exon, mutually exclusive exons, alternative 5' splice site, alternative 3' splice site, intron retention, Composite patterns
JUM does not depend on any prior knowledge of a transcriptome annotation nor known libraries of AS events.
The JUM method is published: https://www.ncbi.nlm.nih.gov/pubmed/30104386 The simulated RNA-seq datasets and ground truth (true AS genes and non-AS genes) described in the paper are deposited in GEO with accession number GSE118193.
Feel free to post any questions or comments on the Google group for JUM users: https://groups.google.com/forum/#!forum/jum-users

JUM