The ABI format is organized with an index (all in one place in the file)
containing offsets to the various pieces of data, along with tags specifying
which piece of data is which. Byte #26 (long int) tells where the index
begins. The data itself is grouped into units by type (sequence base calls,
traces, comments, etc), which can be obtained by reading the offsets from the
index.
A very nice and extremely straightforward run-down of the core data in an ABI
file can be found in David Klatte's technical letter for his ABIView program
(he reverse engineered much of the format):
http://www.users.cloud9.net/~dhk/abitechnical.html
(from his ABIVew homepage http://www.users.cloud9.net/~dhk/abiview.html)
One important note: If you are on an Intel machine, you have to turn all of
the bytes around for the various numerical data. As far as I know, all ABI
(and SCF files, for that matter) are written using Motorola (raw) byte order.
Intel for some reason uses a reverse order.
Hope this helps.
Best regards,
Tom Hall
Hans Engkvist wrote:
> Hello,
> someone here who knows?
> I would like to do my own analysis of the data in a ABI sequencing gel file.
> Unfortuantely I can't figure out in what format/file type the data is in.
> Is it a image format? Other kind of matrix data?
> TIA,
> Hans Engkvist
>> --
> /Hans Engkvist
>> ...before e-mail replying make necessary changes in my e-mail adress...
>> "Nothing shocks me, I'm a scientist"
> /Indiana Jones
