km wrote: > Hi, > why are u reinventing the wheel when Biopython[1] is already existing ? is > there any specific reason u wanted to develop this CoreBio ? why dont u just > extend the existing BioPython package itself ? > regards, > KM > [1]http://biopython.org > <[EMAIL PROTECTED]> wrote:
Biopython is a fine project which I have used and contributed to in the past. Unfortunately, Biopython suffers form a lack of focus. It is a huge heap of code, some of which is well written, but of lot of which is not, and a significant portion doesn't work as advertised. There is no consistency, the code-base is idiosyncratic, the documentation is spotty, and it is very hard to actually discover and use the functionality that you need. For some common tasks there are three ways of doing things, only one of which is supported. CoreBio is intended to be a high-quality, easy to use collection of the core functionality needed for bioinformatics and computational biology. Compared to biopython, we are taking a narrow, quality first, rather than breadth first, approach with simple API's that hide as much complexity as reasonable possible. As an example of simple API's consider the common task of reading a file of protein sequences: >>> from corebio import seq_io >>> afile = open("human.fa") >>> list_of_sequences = seq_io.read(afile) CoreBio will figure out the file format, so 'seq_io.read()' will parse sequence data from fasta, clustal, genbank, intelligenetics, msf, nbrf/pir, nexus or phylip formatted files. Gavin Crooks -- http://mail.python.org/mailman/listinfo/python-list