Collapse transcriptomic and genomic mapping location into a single genomic coordinates

Question

[heyao]

I am using a RNA-seq based fusion detection tools named PRADA (http://www.ncbi.nlm.nih.gov/pubmed/24695405)

The tool uses a modified bwa align the reads to the reference genome and transcriptome at the same time and output all the alignments a read contains.

Next it a customised jar utility map transcript placements to genomic coordinates according to strategies described in this paper, (http://genome.cshlp.org/content/20/4/413.full.html) (see the Supplementary Figure S1)

I just wondering how to implement this idea by code. I think there are two issues:

1. How to map a ensembl transcript(ENST) position back genome coordination ? For example ENST******** contains 3 exons, how to map them back to genome
2. How to convert transcript postion to genome position efficiently. For example a read aligned in ENST********(length 100) position 34, how to identify which exon this read in and map back to genomic coordination?