question about type gapany

[pmr at ebi.ac.uk]

Dear Guy,

> I just noticed a strange thing. The ACD file for emma contains :
>
>   seqall: sequence  [
>     parameter: "Y"
>     type: "gapany"
>   ]
>
> I tried to replace "gapany" by "gap", so that the input sequences should
> be degapped before the alignment procedure starts. I was surprised to get
> an output file with unaligned sequences. As if the degapping was also
> performed on the output instead of just on the input. Is this normal ?
> I ask the question because I have developed programs that work like emma
> an I was busy adding the "type" attributes, in order to improve them. Has
> anybody already seen such behaviour ? Or am I misunderstanding something ?

There is a general principle that the output sequence type matches the
input unless explicitly stated (transeq for example).

This is so that wrappers, GUIs, etc. know what output sequence type to
expect.

You can try to specify gapany as the output sequence type. I think this
will fix your modified ACD file (I'm away from EBI until next week and
will look more closely then).

regards,

Peter