Acquiring very large nucleotide sequence in matlab workspace.

I have been facing the following problems while trying to import very large amt of data in the form of a sequence(~MBases). 1.Using getgenbank does not load the sequence information and CDS headers into Matlab so Iam unable to access some sequences.(Matlab version r2009a) 2. If i use seqtool for a very large sequence eg 'NC_00091', Matlab stops responding. 3. If somehow I am able to access very very large sequences into workspace for some of the organisms using genbank or seqtool, while trying to run the code, matlab exits automatically. And sometimes it shows an error saying data is too large. 4. Do I need to use a bioindexed file for this purpose? Or am I missing out on the memory management part?

[the cyclist]

Can you give a sense of how large "very large" is? How many numbers would you need to store, and what is the nature of those numbers (e.g. are they small integers, or double-precision, etc)?

[Shiwani]

The sequence is an array of the order 1x4639675 nucleotide bases and requires 9279350bytes of mem and data is of type char (nucleotide sequences AGCT...).

Answer 1

[Diego]

What are you trying to accomplish?

The first I'd do is to watch this webinar: Large Data Sets in MATLAB

Other options for handling contig sequences would be:

• -. kent's source tree tools
• -. bioperl

Just my 2 cents,

Regards,

Diego