On Freitag 04 Dezember 2009 Jeremiah Davie wrote:
Hi Everyone,
Quick question: In my assembly of 454 data (older GS20 data from
standard chemistry), my output log lists that it found 290768 reads
extracted from two *.sff files, yet when those reads are loaded I
receive 145339 instances of the following error message: "Error: read
name EERNOEK******* present multiple times in readpool!", where the
asterisks denote a different read name.
Hello Jeremiah,
well, if it's not a bug in MIRA (and I don't think it is), then you
really
have sequences named identically in the two SFF files. Whether the
sequences
themselves are identical remains to be checked (and you should do
that)
Since this is almost half of
the the total loaded reads, I'm curious what might have caused this?
Presumably sff_extract used the names found within the *.sff files
for
these reads, but it seems strange to me that the two *.sff files
would
contain all the same names.
I think I remember reading in some Roche documentation that they
changed the
read naming somewhen to ensure more diversity or something similar.
On the other hand, it's absolutely possible to recombine SFF files
into
different SFF files with the sff-tools from the Roche pipeline, so
that may be
an explanation if you have identical reads in several SFF files. Or
maybe one
SFF has 'clipped' versions of reads from the other SFF.
Did I do some thing wrong in the
sff_extract process that might have caused this without generating an
error message?
No, you didn't do anything wrong. sff_extract (when used on unpaired
data)
takes the read names verbatim from the SFF file, so they're in there
already.
(on paired-end data, there's some name mangling, but that consists of
appending postfixes, so no way something could go wrong there neither)
Any thoughts? Thanks in advance everyone! - Jeremiah
Qickest thing to do: take a few examples where MIRA complained about
double
read names and check the sequences by eye in the FASTA file. You
should be
able to decide pretty quickly whether the sequences are the same or
perhaps a
clipped subset of each other. Then you'll need to decide which
version you
want to take.
In the most improbable case that the reads have identical names but
the
sequences have nothing in common ... well, you would need to massage
the read
names a bit. For that, I would extract both SFF separately, put a
prefix for
each read name of each FASTA, FASTA quality and XML file (that's one
"sed"
command :-) and then put everything back together into one FASTA,
one FASTA
quality and one XML file.
Hope that helps, and I'd be curious to know what it turned out to be.
Regards,
Bastien
--
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
