[Bioperl-l] BPslite and iterations

Peter Schattner schattner@alum.mit.edu
Fri, 30 Mar 2001 08:39:50 -0800


Arne Mueller wrote:

>
> this might be very trivial but probably useful, too ;-) . The BPslite
> class could have a method 'converged' to determine whether the whole
> psiblast run has been converged (i.g. the 'CONVERGED' string was found).
> The obvious workaround at the moment is to compare the number of maximum
> iterations (-j) with $report->number_of_iterations, but this cannot be
> done if the model converges in e.g. iteration 20 when -j was set to 20.

Sounds reasonable.  I'll plan to add such a method to bppsilite before the
next release.  On the other hand, if you need it quicker, feel free to
write one yourself and add it to the repository ;-)

>
>
> Also, are there any future plans to parse xml output from blast/psiblast
> ?

Not that I am aware of.

Regards

Peter