Home

About BRIE

BRIE (Bayesian regression for isoform estimate) is a Bayesian method to estimate isoform proportions from RNA-seq data. Currently, BRIE could take sequence features to automatically learn informative prior of exon inclusion ratio in exon-skippiing events. This informative prior is very important when limited data is available. In Bulk RNA-seq experiment, we could easily increase the amplification to get more sequencing reads to improve the accuracy of isoform estimate. However, in single cell RNA-seq (scRNA-seq) experiments, the initial molecular is very limited, which always results some genes with very low coverage or even drop-out. In scRNA-seq, the BRIE method, by integrating informative prior, e.g. learned from sequence feature, could provide accurate and reproducible estimates of splicing in single cells, as well as sensitive differential analyses.

BRIE provides following functions through command line:

1. brie: Estimate isoform proportions and FPKM, and calculate weights for regulatory features.

2. brie-diff: Calculate Bayes factor of differential splicing between multiple cells pair-wisely.

References

Yuanhua Huang and Guido Sanguinetti. BRIE: transcriptome-wide splicing quantification in single cells. Genome Biology, 2017; 18(1):123.