# so far the following merge seems to be 100% sufficient for the desired final output ...
# we could add on extra features if needed ...
quant_n_raw = quant_info_unrolled[['pept',
'deamid_info',
'pept_with_mod',
'Weight',
'spec_name',
'Mascot Ion Score',
'Mascot Identity Score',
'Mascot Delta Ion Score',
# 'prot_ident_probab', # in case of ambiguity we need more columns to merge on !!!
'pept_ident_probab']].merge(bad_info,how='right',on=['pept','spec_name'],suffixes=('','_x'))
#######################################################
# Now, extract those gsites ...
dg_func = lambda x: pd.Series( ms.deamid_to_gsite(x['deamid_info'], x['start_fetched'], str(gbrecs[str(int(x['fetchid']))].seq)) )
# and add them back to the main table ...
gs_res = quant_n_raw[['deamid_info','start_fetched','fetchid']].apply( dg_func, axis=1 )
quant_n_raw = quant_n_raw.merge(gs_res,left_index=True,right_index=True)
print
print "Now we'd need to add theoretical glycosilation sites as a separate column ..."
print "full protein sequence and its length is added as well ..."
# this analysis must be done, once for each 'fetchid', and then merged back to the main table ...
get_theor_sites_fid = lambda fid: ms.get_theor_sites(str(gbrecs[str(fid)].seq))
get_theor_sites_number_fid = lambda fid: ms.get_theor_sites_number(str(gbrecs[str(fid)].seq))