3.2 years ago by
From testing various inputs, it seems that this tool has a input line length upper limit. (Likely for technical reason, not a specific setting).
Doing either one of the following resulted in a successful run for your failures.
- Wrap the fasta sequence with the tool FASTA Width formatter with a width between 40-80. 40 is less commonly used, I tend to use 60, but 80 is the default for most. This creates input that is in strict fasta format (a requirement implied by the tool help stating "sequence" input e.g fasta format, with more details in the tool documentation). For these reasons, it is the prefered solution.
- Reduce the length of the sequence identifiers. Run Fasta-to-Tabular with the option "How many title characters to keep?". I set this at 20. Then converted back to fasta with Tabular-to-Fasta. Executing the tool with this input was successful. So, the tool is able to accept non-wrapped fasta data, but it appears that length is a factor for this to work. If you ever have a dataset that fails with the same original error after using this method, go ahead and wrap the lines, as the input line length was likely exceeded again, despite the identifier trimming.
Hope this helps! Jen, Galaxy team