[mira_talk] Re: Quality of merged Solexa reads
- From: Bastien Chevreux <bach@xxxxxxxxxxxx>
- To: mira_talk@xxxxxxxxxxxxx
- Date: Sat, 12 Jun 2010 22:23:58 +0200
On Freitag 11 Juni 2010 Christina Toft wrote:
> [...]
> My question is regarding the quality assigned of the merged solexa
> reads. In the caf file that MIRA produced for this assembly I have 17
> merged reads (_cer_sxa_) which has “normal” quality scores (30-40) but
> the rest of the solexa reads has a quality score of 1. Together the 17
> merged solexa reads span the whole genome and they never overlap each
> other. How come it’s only one solexa base per column that get assigned
> a quality? Is there any way I can get MIRA to output quality scores
> for the other solexa reads as well?
Merged Solexa reads means that the original Solexa reads don't exist anymore.
What MIRA does is this:
- it keeps track, for each position in the reference sequence, two
values:
a) the number of times a perfect, 100% match covered that position
b) the highest quality of that base at that perfect matching position.
- reads without 100% match are kept as stand-alone reads
When writing the results, MIRA will recreate reads (CERs = Coverage Equivalent
Reads) so that the coverage is attained. Using reads of length 4 as example:
Ref: ACGTGTGTACGTAGT ACGTGTGTACGTAGT
oread1 acgt acgtgtgtacgta : cer1
oread2 cgtg cgtgt tagA : cer2 & oread6
oread3 gtgt ====> gt : cer3
oread4 tgta
oread5 cgta
oread6 tagA
In the above example, oread6 has a difference to the reference and is kept
apart. oread1 to oread5 "disappear" into cer1, cer2 and cer3. As only one
quality value has been kept per reference position, only 1 qual value can be
assigned. In the example above, cer1 gets them. cer2 and cer3 get dummy values
of "1".
So, no, if you used merged reads, you can only get 1 quality value per
reference position. But that usually is enough, at least has been for me.
B.
Therefore, no, if you are using merged reads
--
You have received this mail because you are subscribed to the mira_talk mailing
list. For information on how to subscribe or unsubscribe, please visit
http://www.chevreux.org/mira_mailinglists.html
Other related posts:
