RNA-MATE: a recursive mapping strategy for high-throughput RNA-sequencing data

Abstract

Mapping of next-generation sequencing data derived from RNA samples (RNAseq) presents different genome mapping challenges than data derived from DNA. For example, tags that cross exon-junction boundaries will often not map to a reference genome, and the strand specificity of the data needs to be retained. Here we present RNA-MATE, a computational pipeline based on a recursive mapping strategy for placing strand specific RNAseq data onto a reference genome. Maximizing the mappable tags can provide significant savings in the cost of sequencing experiments. This pipeline provides an automatic and integrated way to align color-space sequencing data, collate this information and generate files for examining gene-expression data in a genomic context.

Availability: Executables, source code, and exon-junction libraries are available from http://grimmond.imb.uq.edu.au/RNA-MATE/

Fig. 1.

The RNA-MATE recursive mapping pipeline. The pipeline consists of four major components. (1) Tags are (optionally) filtered based on the quality values for each basecall. (2) The alignment module attempts to align tags first to the genome, and then to a library of known exon-junction sequences. If a tag fails to align, then the tag is truncated, and the process is repeated. (3) The optional tag rescue module uses information derived from both single- and multi-mapping tags to uniquely place multi-mapping tags. (4) Finally, UCSC genome browser compatible wiggle plots and BED files are generated.