Open jwasmuth opened 9 years ago
Would you be able to provide some input data that replicates the problem, as well as the exonerate version (e.g., git commit SHA-1) and arguments used?
Same issue here. I ran:
$ fastaexplode -f LG05.fasta
And I got the following error:
** FATAL ERROR **: Unrecognised symbol '-' (ascii:45) file:[LG05.fasta] seq:[Br] pos:[37388847]
exiting …
The fasta file I used is available at https://github.com/mmatschiner/Introgression-Tutorial/blob/master/LG05.fasta.zip?raw=true
same issue for me
was this ever solved? I am getting the same error and cannot find any error in the input query. Is it possible that it is caused at some point by the 6-frame translation of the genome? some scaffolding error or similar?
Hi there, I'm getting the same error now too - is there a fix for this or any clues as to the root cause?
in case it helps, sometime the problem is caused by the assembly, which should be a line-even fasta.
@dukecomeback sorry for the late reply - what do you mean by line-even fasta?
@katieemelianova One-line fasta refers to those with sequence shown in one-line. So if you less -S it, you will see: >name1 actgactgactg........ >name2 gactgactgactg........
I mean by line-even fasta those with sequence broken into different line. So if you less -S it, you will see: >name1 actgcatgactg acgcatdgcac cccccgcaaaa .... >name2 actgcatgactg acgcatdgcac cccccgcaaaa .....
You can use fastareformat (in exonerate toolkit) to transfer one-line fasta into line-even fasta. Sorry if I brought the wrong nomenclature.
@dukecomeback thanks for the clarification, sorry for the misunderstanding! :)
I actually checked and I dont think that it is a newlines problem, I took the contig that I think was the problem and ran it with both newlines and without, and it seemed to work fine both ways :/ so either I have tested the wrong contig or there is some glitch in exonerate which only happens occasionally.
When "--showalignment yes" I continually get "\ FATAL ERROR **: Unknown amino acid [-]". When "--showalignment no" it's fine. Is this a problem with the gap symbol? Thanks