Appropriate Reference Format

[gaffa]

I'm not an expert on exome sequencing but from what I gather you should probably always align reads to the whole genome, not only to the regions on your capture array. However this is for general conceptual reasons, I don't know what's going on with your particular files.

There has been some discussion of this on Biostar:

http://biostar.stackexchange.com/questions/7572/is-there-any-reference-exome/7617#7617

http://biostar.stackexchange.com/questions/4413

[HSG_student1]

Hello,

I am trying to deal with some targeted resequencing data. I have a subset of hg19 that I targeted with a SureSelect Capture array and now have some Illumina HiSeq data on it. I aligned the reads to my .fasta formatted reference with Novoalign. The reference was downloaded from UCSC as exonic regions of known transcripts. An example is below. I am not aligning to all of h19 because I just want to see the alignments to the targeted regions. Do I need to align to the whole genome instead?

>hg19_ct_UserTrack_3545_hg19_knownGene_uc001cyp.2_8 range=chr1:57409309-57409587 5'pad=30 3'pad=30 strand=- repeatMasking=none
AACAGGGCTAGGCCTCTGGACTCGATCATAGGAAAAGGGACAGAGTGAAC
TACCTTTTACGTCATGGACCTTTTTCCTTGTTTTCTTCAGATTATACTCT
TAACAACGTCCATGCCTGTGCCAAAAATGATTTTAAAATTGGTGGTGCCA
TTGAAGAGGTCTACGTCAGTCTGGGTGTGTCTGTAGGCAAATGCAGAGGT
ATTCTGAATGAAATAAAAGGTGAGTGACAGGGGCTTGTGACTTCCACACT
TATAATGTTTATCCTCCTAGTGCCTCTGG
>hg19_ct_UserTrack_3545_hg19_knownGene_uc001aru.2_11 range=chr1:11086521-11087795 5'pad=30 3'pad=30 strand=- repeatMasking=none
TCAAATATATACTTCAAGGCGGTCATAATTTAAGAAAAACACGATTGACA
TTTTTACAAAGGCTCTCTACCTATCTGCTTCTTTCTCCAGTTTGTGGACT
ATCAGCCCGCACAACAGGAGGGCGTATATATGGAGGGCAAAAGGCAAAAC

The reads align fine using the above reference file in NovoAlign and I was able to make a .bam file with samtools and sort it in anticipation of making a bedGraph or Wiggle file using BEDTools. However, the headers used in my reference .fasta file do not seem to be compatible with the hg19.genome file that comes with BEDTools (a few lines are posted here):

chr1 249250621
chr2 243199373
chr3 198022430
chr4 191154276.

Any advice if I should redownload my targeted reference sequence in a different format? Or parse to contain some different header format, for example:
>chr1 start:xxxxxxx end:xxxxxx

I have found lots of great help for ways to manipulate the files, but can't seem to conquer this most basic step.

Thanks alot,
Mike