transcripts-w-groupBy.sh

# Step 1: Get transcripts from UCSC refGene (hg19) into a BED file.
# Notes:
#      the awk statement reorders the "raw" columns into BED12 format
#      bed12ToBed6 converts the BED12 into discrete BED6 entries for each exon
#           - the -n option is new and in the bedtools repository
$ curl -s http://hgdownload.cse.ucsc.edu/goldenPath/hg19/database/refGene.txt.gz | \
      zcat | \
      awk '{OFS="\t"; print $3,$5,$6,$2,$9,$4,$7,$8,"0",$9,$10,$11}' | \
      bed12ToBed6 -n \
      > refGene.bed

$ head refGene.bed
chr19	50595745	50595866	NR_024227	2	-
chr19	50601082	50601203	NR_024227	2	-
chr16	5464988	8197482	NM_018992	1	+
chr16	5478429	8224364	NM_018992	2	+
chr16	5480400	8228306	NM_018992	3	+
chr16	5482315	8232136	NM_018992	4	+
chr16	5484847	8237200	NM_018992	5	+
chr16	5489792	8247090	NM_018992	6	+
chr12	237002794	355504191	NM_019086	10	-
chr12	237007578	355513759	NM_019086	9	-


# Step 2: Use fastaFromBed to extract the sequence for each exon