On 08 Oct 2014, at 18:01 , Magnus Popp <magnus.popp@xxxxxxxxxx> wrote: > Concerning your examples below, I’m OK with the first case as I would > interpret an IUPAC "R" as a result of merging what probably are two alleles, > but ideally mira would give me two contigs. I see that wasn’t what I wrote in > my question, but that’s what I meant, sorry… Here again, MIRA will give you two contigs if there is enough evidence for each base. Getting IUPAC bases means: there was not enough evidence to discard a sequencing error as possibility, but there is too much evidence to make make a clean base call. > In second example the preferred result would be that the contig (or perhaps > the read) is clipped when the quality drops that low. > > I will align each RAD loci/“EST” across a handfull of fairly closely related > species and although the errors (be it a single base or a IUPAC base) most of > the time only will result in apomorphic characters, they will mess up the > diversity and branch length estimates in my phylogenetic analyses. So I > rather have somewhat shorter RAD loci/ESTs with ± reliable data than longer > ones with shaky 3’ ends at this stage. If push comes to shove, use the option to force non-IUPAC base calling: -CO:fnicpst Well, forcing may be a bit too much said: it will reduce the IUPACs. > Unless you suggest otherwise, I’ll probably use miraconvert and set -q to > something that works with my data and then simply truncate at the first N or > perhaps just merge the fasta and qual files and use a suitable cut off in > fastq quality trimmer. Well, I just looked up the code and found the reason why I deactivated it: a class in command line parameters. It used to be “-v”, but I since moved the version output to this. I simply forgot to reassign a new parameter. “-V” seems to be free … I’ll have a look whether I can give you a snapshot of the dev tree later-on this week. B. -- 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