Recently I need to generate genome index files for Ion Torrent results including ERCC, so I just used Gencode sequences and annotations.

Fasta files were downloaded here, it is hg19 with patch version 13, and v19 is the last version of Gencode based on hg19. GTF annotation is this.

ERCC sequences and annotation files were downloaded from LifeTech.

pyfasta was used to split the fasta file into small chromosome-based fasta files.

## Combine the sequences and annotations
cat GRCh37.p13.genome.fa ERCC92.fa > hg19_P13_ERCC.fa
cat gencode.v19.annotation.gtf ERCC92.gtf > hg19_P13_ERCC.gtf

## Split the fasta file into chromosome-based fasta files
## 1. for MapSplice
## 2. to modify the header of chromosomes and contigs
pyfasta split --header "%(seqid)s.fa" hg19_P13_ERCC.fa
rm hg19_P13_ERCC.fa

## Remove haplotype files we don't use
rm GL*.fa
rm JH*.fa
rm K*.fa

## Remove space and suffix in file name.
## Quotes were used in "${i}" because the files contain space.
for i in chr*.fa; do
	mv "${i}" $(echo $i | sed -e 's/ [0-9A-Z]*//g')
done

## Modify fasta header
for i in chr*.fa; do
	sed -i -e '1s/ [0-9A-Z]*//g' ${i} 
done

cat chr*.fa ERCC*.fa > hg19_P13_ERCC.fa

Now hg19_P13_ERCC.fa and hg19_P13_ERCC.fa are ready for the buiding of index.

## Build index for STAR
STAR --runMode genomeGenerate --genomeDir ./hg19_gencodev19_ERCC --genomeFastaFiles hg19_P13_ERCC.fa --runThreadN 8 -sjdbGTFfile hg19_P13_ERCC.gtf -sjdbOverhang 100

bowtie-build hg19_P13_ERCC.fa hg19_P13_ERCC
bowtie2-build hg19_P13_ERCC.fa hg19_P13_ERCC