[feature] Add support for gtf in load_annotation()

[shouldsee]

GTF is so widely used. Should be easily addable with BCBio.GFF

import Bio

from BCBio.GFF import GFFExaminer
import BCBio.GFF as bgff

import pymisca.util as pyutil
import pymisca.vis_util as pyvis
plt = pyvis.plt

def read_gtf(in_file, cache=1):
    if isinstance(in_file,str):
        in_handle = open(in_file)    
    else:
        in_handle = in_file
    it = bgff.parse(in_handle,)
    if cache:
        res = list( it )
        in_handle.close()
    else:
        res = it
    return res



def gene2transcript(g,force=0):
    '''Convert a SeqRecord from BCBio.GFF.parse() to a dictionary-like object
'''
    if isinstance(g,Bio.SeqRecord.SeqRecord):
        if not force:
            assert len(g.features)==1
        g = g.features[0]
        
    feats = g.sub_features
    d = {'parent':g}
    for i,f in enumerate(feats):
        if f.type in ['start_codon','stop_codon']:
            d[f.type]  = f
    return pyutil.util_obj(**d)

[simonvh]

Sorry! This took a while, as I was very busy. GTF/GFF is indeed widely used, however, especially for GFF there are many dialects that are subtly different. But I'll put it on the list of future enhancements.