[Biopython] NCBIWWW and megablast

[Peter]

On Fri, Oct 8, 2010 at 2:34 PM, Bart <bratdaking at gmail.com> wrote:
> Hi Peter,
>
> hereby the patch.

Thank's Bart.

I'll make some minor changes - in particular if you add a new optional
argument to an existing function, it should be at the end. This is for
backwards compatibility in case anyone was supplying their arguments
by order. e.g. If you had this:

def f(a, b=None, c=True):
   ...

answer = f("test", None, False)

If you added a fourth argument with 'def f(a, b=None, c=True, d=10)'
then the above example is unaffected. However, if you insert the new
argument earlier, e.g. 'def f(a, b=None, d=10, c=True)'  the example
breaks.

> I found another argument missing: PSSM. It is to add a PSI
> BLAST checkpoint. No clue how this should be done precisely,
> but I added the lines anyway.

Curious - last time we looked at this, the NCBI didn't seem to support
PSI-BLAST via the online API,
http://bugzilla.open-bio.org/show_bug.cgi?id=2496

> The third argument missing is the RESULTS_FILE parameter. I have
> left that one out as I assume that that one needs somewhat more
> alterations to be able to also download the file.
>
> Cheers,
> Bart

I think we should just put the PSSM and RESULTS_FILE parameters
into the code as comments (until we know how and if we can use them).

Peter