[Biopython-dev] SeqIO.InterlacedSequenceIterator

[Peter Cock]

On Fri, Dec 14, 2012 at 10:07 AM, Lucas Sinclair <lucas.sinclair at me.com> wrote:
> Hello,
>
> Thanks for your response. Yes I looked at Bio.SeqIO.index, it makes
> an index, but it is held in memory. So it must be recomputed every
> time the interpreter is reloaded.

Yes, that is right.

> This step is wasting enough time for me that I would like to compute
> the index on my 50GB file once, and then be done with it. SQLite
> really is the technology of choice for such a problem...

Yes, which is why Bio.SeqIO.index_db() stores the index in SQLite.
The SeqIO chapter in the Tutorial does try to explain this and the
advantages compared to Bio.SeqIO.index(). Have you tried this yet?

> I suppose you agree storing all this sequence information in flat
> ascii files is not piratical.

It may not be optimal, but it is very practical (although at the scale
of next generation sequencing data less so).

Peter