Merging different reference genomes

[g.galazzo]

Hi to everyone,
i'm trying to replicate a pipeline from an article in which they says that they had merged different information from different reference genomes to build an unique and complete reference genome.
The article is by Billy Li and here is the link

http://www.nature.com/nmeth/journal/v9/n6/pdf/nmeth.1982.pdf

and what they said is : " we mapped RNA-seq
reads not only to the reference genome or to the transcriptome
but to a combination of the hg19 reference genome plus exonic
sequences surrounding all currently known splicing junctions
from gene models available in annotation from Gencode, RefSeq,
Ensembl and UCSC Genes"
So here is my problem, i don't understand how they performed this task.
I tried in different ways to merge the reference genomes,
i tried merging the genome.fa file from Wholegenome folder from my reference files,
i tried to merge the chromosome files together and after fuse them in a unique file,
i also tried to merge the genome.fa from the BWAIndex folder that is already present in to the reference folders, but nothing works.
I index the merged files but the program doesn't recognize the index.
does anyone have and idea about how to perform this merging?

thanks all

[dpryan]

Honestly, it would make more sense to just use a split-read mapper like STAR, tophat, or segemehl. You can then just map to the genome and be done with that step instead of needing to play around with adding splice junction sequences in.