[ensembl-dev] Wiggletools bam coords and coverage mode

Michael Nuhn mnuhn at ebi.ac.uk
Fri May 5 14:13:59 BST 2017


Hello Nathan,

the cigar string from the first read shows that the first 49 bases at 
the beginning are soft clipped (49S).

The part that actually matches begins at base 10_049, not 10_000.

Wiggletools ignores the clipped bases. That is why wiggletools shows you 
one read, but not the other.

Cheers,
Michael.


On 05/05/17 12:09, Nathan Johnson wrote:
> Morning
>
> I'm just fiddling with the coverage mode and saw what looks like a 1/2
> open issue, or maybe an assumption that bams are using 1/2 open coords.
>
> Samtools show 2 reads starting at 10000:
> ```
> $ samtools view test.bam chr1:1-10000
> NB501627:84:HMJNWBGX2:3:22601:21701:15135    121    chr1    10000    0
>   49S30M    =    100000
> AATAAATTATAGATTATATAAAAATTAATTAATAATGCTTCCCCTATCCATAACCCTAACCCTAACCCTAACCCTAACC
>
>   //AE//////////EE////////////E///E/////E/////A/////E//////EEEEEAEEEEEEEEEEEAAAAA
> NB501627:84:HMJNWBGX2:3:22601:21701:15135    181    chr1    10000    0
>   *    =    100000
> AATATTAATTAATTATTATTTAATCGATAATATATTAATATTATTTAATCGTATTTAAATTGTTTGTGGTTGTGGGTGAG
>
>   EEEE/EEAAEEEE/EEAEEEEEEEEEEEEEEEEEEEEEEEEEEEEAEEEEEAE/<EEE/EEEEEEEEEA/EEEEEA/AAA
> ```
> Tags truncated for brevity.
>
> Wiggletools shows none:
> `$ wiggletools seek chr1 1 10000  coverage test.bam`
>
> When extending to 10001, the coverage value only show 1, not 2?
> ```
> $ wiggletools seek chr1 1 10001  coverage test.bam
> fixedStep chrom=chr1 start=10000 step=1
> 1.000000
> ```
>
> Can someone confirm what's going on here?
>
> Thanks
>
> Nathan
>
>
> _______________________________________________
> Dev mailing list    Dev at ensembl.org
> Posting guidelines and subscribe/unsubscribe info: http://lists.ensembl.org/mailman/listinfo/dev
> Ensembl Blog: http://www.ensembl.info/
>



More information about the Dev mailing list