Some future ideas

[Bernd Jagla]

David Martin wrote:

> I had som ethoughts about future expansions of EMBOSS and really want to
> put down a few 'placemarkers' to see what people think.
>
>         Eventually people are going to get the idea that EMBOSS should be
> able to do things with data other than just sequences and want to run
> e.g. microarray and structure type analyses. With the current database
> definitions in emboss.default there is a type: clause that is not
> required. I would propose to make this mandatory and extend the values.
>
> N is nucleotide sequence database
> P is protein sequence database
> S is a structure database
> M is a microarray experiment database.
>
> The USA can then be extended to cover structures (pdb:1HTF) for an example
> and microarray experiments. There are probably other entities that could
> be included.
>
> With some careful type management we could even convert types on the fly,
> so you could put in a pdb reference when asked for a protein sequence and
> it would be automatically derived (OK, there are a lot of problems with
> such things but it would be useful).
>
> Other possibilities:
>
> XML format output in some suitable XML format? This would probably need
> a lot of work in the libraries to tidy everything up and make it work.
>
> Still looking for a student to write an EMBOSS-WAP interface ;-)
>
> ..d
>
> ---------------------------------------------------------------------
> *  Dr. David Martin                  Biotechnology Centre of Oslo   *
> *  Node Manager                      Gaustadalleen 21               *
> *  The Norwegian EMBNet Node         P.O. box 1125 Blindern         *
> *  tel +47 22 95 87 56               N-0317 Oslo                    *
> *  fax +47 22 69 41 30               Norway                         *
> ---------------------------------------------------------------------

Hi David,

I still feel quite new in EMBOSS and am not that familiar with the databanks,
but it sounds very good to be able to analyze some micro array data.

I also believe that there should be some other possibilities for data
analysis. I personally like artificial neural network for they are fast,
"easy" to use and I have already some programs written using EMBOSS. I am
thinking of some other statistical analysis tools to implement (information
analysis, some visual output of aa content, distribution and so many other
things). For this it would be a good to be able to build groups of sequences
and sequence parts, add some numbers to these groups, have probably a new
class of functions dealing with these groups.

Of course, we should discuss the data model a little more in detail if it is
interesting...

So, do you thing EMBOSS should be able to deal with these  kind of problems
as well?

Bernd