Bio.FSSP

1 """Parser for FSSP files, used in a database of protein fold classifications. 2 3 This is a module to handle FSSP files. For now it parses only the header, 4 summary and alignment sections. 5 6 See: Holm and Sander (1996) The FSSP database: fold classification based on 7 structure-structure alignment of proteins. 8 9 functions: read_fssp(file_handle): reads an fssp file into the records. Returns a 10 tuple of two instances. 11 mult_align: returns a Biopython alignment object 12 """ 13 import re 14 import fssp_rec 15 from Bio.Align import Generic 16 from Bio import Alphabet 17 fff_rec = fssp_rec.fff_rec 18 header_records = { 19 'database': re.compile('^DATABASE'), 20 'pdbid': re.compile('^PDBID'), 21 'header': re.compile('^HEADER'), 22 'compnd': re.compile('^COMPND'), 23 'author': re.compile('^AUTHOR'), 24 'source': re.compile('^SOURCE'), 25 'seqlength': re.compile('^SEQLENGTH'), 26 'nalign': re.compile('^NALIGN') 27 } 28 29 summary_title = re.compile('## +SUMMARY') 30 summary_rec = re.compile(' *[0-9]+: +[1-9][0-9a-z]{3,3}') 31 alignments_title= re.compile('## +ALIGNMENTS') 32 alignments_rec = re.compile(' *[0-9]+ +-{0,1}[0-9]+') 33 equiv_title = re.compile('## +EQUIVALENCES') 34 35

36 -class FSSPHeader(object):

37 - def __init__(self):

38 self.database = None 39 self.pdbid = '' 40 self.header = '' 41 self.compnd = '' 42 self.source = '' 43 self.author = [] 44 self.seqlength = 0 45 self.nalign = 0

46

47 - def fill_header(self, inline):

48 for i in header_records: 49 if header_records[i].match(inline): 50 if i == 'database' or i == 'seqlength' or i == 'nalign': 51 setattr(self, i, int(inline.split()[1])) 52 elif i == 'compnd' or i == 'author': 53 setattr(self, i, inline.split()[1:]) 54 elif i == 'source' or i == 'header': 55 attr = inline[inline.find(' ')+1:].strip() 56 setattr(self, i, attr) 57 else: 58 setattr(self, i, inline.split()[1])

59 60

61 -class PosAlign(object):

62 - def __init__(self, inStr):

63 inStr = inStr.strip() 64 if len(inStr) != 1 and len(inStr) != 2: 65 raise ValueError('PosAlign: length not 2 chars' + inStr) 66 if inStr == '..': 67 self.aa = '-' 68 self.gap = 1 69 else: 70 self.gap = 0 71 self.aa = inStr[0] 72 if self.aa == self.aa.lower(): 73 self.aa = 'C' 74 if len(inStr) == 2: 75 self.ss = inStr[1].upper() 76 else: 77 self.ss = '0'

78

79 - def __repr__(self):

80 if self.gap: 81 outstring = '..' 82 else: 83 outstring = self.aa+self.ss.lower() 84 return outstring

85 86 __str__ = __repr__

87 88

89 -class FSSPSumRec(object):

90 """ Contains info from an FSSP summary record"""

91 - def __init__(self, in_str):

92 self.raw = in_str 93 in_rec = in_str.strip().split() 94 # print in_rec 95 self.nr = int(in_rec[0][:-1]) 96 self.pdb1 = in_rec[1][:4] 97 if len(in_rec[1]) == 4: 98 self.chain1='0' 99 elif len(in_rec[1]) == 5: 100 self.chain1=in_rec[1][4] 101 else: 102 raise ValueError('Bad PDB ID 1') 103 self.pdb2 = in_rec[2][:4] 104 if len(in_rec[2]) == 4: 105 self.chain2='0' 106 elif len(in_rec[2]) == 5: 107 self.chain2=in_rec[2][4] 108 else: 109 raise ValueError('Bad PDB ID 2') 110 self.zscore = float(in_rec[3]) 111 self.rmsd = float(in_rec[4]) 112 self.lali = float(in_rec[5]) 113 self.lseq2 = float(in_rec[6]) 114 self.pID = float(in_rec[7]) 115 self.revers = int(in_rec[8]) 116 self.permut = int(in_rec[9]) 117 self.nfrag = int(in_rec[10]) 118 self.topo = in_rec[11] 119 self.doc = '' 120 for i in in_rec[12:]: 121 self.doc = self.doc + i + ' ' 122 self.doc = self.doc.rstrip() + '\n'

123

124 - def __repr__(self):

125 return self.raw

126 __str__ = __repr__

127 128

129 -class FSSPAlignRec(object):

130 - def __init__(self, in_fff_rec):

131 # print in_fff_rec 132 self.abs_res_num = int(in_fff_rec[fssp_rec.align.abs_res_num]) 133 self.pdb_res_num = in_fff_rec[fssp_rec.align.pdb_res_num].strip() 134 self.chain_id = in_fff_rec[fssp_rec.align.chain_id] 135 if self.chain_id == ' ': 136 self.chain_id = '0' 137 self.res_name = in_fff_rec[fssp_rec.align.res_name] 138 if self.res_name == self.res_name.lower(): 139 self.res_name = 'C' 140 self.ss1 = in_fff_rec[fssp_rec.align.ss1] 141 self.turn3 = in_fff_rec[fssp_rec.align.turn3] 142 self.turn4 = in_fff_rec[fssp_rec.align.turn4] 143 self.turn5 = in_fff_rec[fssp_rec.align.turn5] 144 self.pos_align_dict = {} 145 self.PosAlignList = []

146

147 - def add_align_list(self, align_list):

148 for i in align_list: 149 self.PosAlignList.append(PosAlign(i))

150

151 - def pos_align_list2dict(self):

152 j = 1 153 for i in self.PosAlignList: 154 self.pos_align_dict[j] = i 155 j = j + 1

156 157

158 -class FSSPAlignDict(dict):

159 - def __init__(self):

160 # The following two dictionaries are pointers to records in self 161 # The first dictionary is a "pdb_residue_number: self_key" 162 # The second dictionary is a "absolute_residue_number: self_key" 163 self.pdb_res_dict = {} 164 self.abs_res_dict = {} 165 self.data = {}

166

167 - def build_resnum_list(self):

168 for i in self: 169 self.abs_res_dict[self[i].abs_res_num] = i 170 self.pdb_res_dict[self[i].pdb_res_num] = i

171 172 # Given an absolute residue number & chain, returns the relevant fssp 173 # record

174 - def abs(self, num):

175 return self[self.abs_res_dict[num]]

176 177 # Given an PDB residue number & chain, returns the relevant fssp 178 # record

179 - def pdb(self, num):

180 return self[self.pdb_res_dict[num]]

181 182 # Returns a sequence string

183 - def sequence(self, num):

184 s = '' 185 sorted_pos_nums = self.abs_res_dict.keys() 186 sorted_pos_nums.sort() 187 for i in sorted_pos_nums: 188 s += self.abs(i).pos_align_dict[num].aa 189 return s

190

191 - def fasta_mult_align(self):

192 mult_align_dict = {} 193 for j in self.abs(1).pos_align_dict: 194 mult_align_dict[j] = '' 195 for fssp_rec in self.itervalues(): 196 for j in fssp_rec.pos_align_dict: 197 mult_align_dict[j] += fssp_rec.pos_align_dict[j].aa 198 seq_order = mult_align_dict.keys() 199 seq_order.sort() 200 out_str = '' 201 for i in seq_order: 202 out_str += '> %d\n' % i 203 k = 0 204 for j in mult_align_dict[i]: 205 k += 1 206 if k % 72 == 0: 207 out_str += '\n' 208 out_str += j 209 out_str += '\n' 210 return out_str

211 212

213 -class FSSPSumDict(dict):

214 pass

215 216 217 # 218 # Process a fssp file into its constituents. Return a 2-tuple containing 219 # a list of FSSPSumRecs and a dictionary of alignment records. 220 #

221 -def read_fssp(fssp_handle):

222 header = FSSPHeader() 223 sum_dict = FSSPSumDict() 224 align_dict = FSSPAlignDict() 225 # fssp_handle=open(fssp_handlename) 226 curline = fssp_handle.readline() 227 while not summary_title.match(curline): 228 # Still in title 229 header.fill_header(curline) 230 curline = fssp_handle.readline() 231 232 if not summary_title.match(curline): 233 raise ValueError('Bad FSSP file: no summary record found') 234 curline = fssp_handle.readline() # Read the title line, discard 235 curline = fssp_handle.readline() # Read the next line 236 # Process the summary records into a list 237 while summary_rec.match(curline): 238 cur_sum_rec = FSSPSumRec(curline) 239 sum_dict[cur_sum_rec.nr] = cur_sum_rec 240 curline = fssp_handle.readline() 241 242 # Outer loop: process everything up to the EQUIVALENCES title record 243 while not equiv_title.match(curline): 244 while (not alignments_title.match(curline) and 245 not equiv_title.match(curline)): 246 curline = fssp_handle.readline() 247 if not alignments_title.match(curline): 248 if equiv_title.match(curline): 249 # print "Reached equiv_title" 250 break 251 else: 252 raise ValueError('Bad FSSP file: no alignments title record found') 253 254 if equiv_title.match(curline): 255 break 256 # If we got to this point, this means that we have matched an 257 # alignments title. Parse the alignment records in a loop. 258 curline = fssp_handle.readline() # Read the title line, discard 259 curline = fssp_handle.readline() # Read the next line 260 while alignments_rec.match(curline): 261 align_rec = FSSPAlignRec(fff_rec(curline)) 262 key = align_rec.chain_id + align_rec.res_name + str(align_rec.pdb_res_num) 263 align_list = curline[fssp_rec.align.start_aa_list:].strip().split() 264 if key not in align_dict: 265 align_dict[key] = align_rec 266 align_dict[key].add_align_list(align_list) 267 curline = fssp_handle.readline() 268 if not curline: 269 print 'EOFEOFEOF' 270 raise EOFError 271 for i in align_dict.itervalues(): 272 i.pos_align_list2dict() 273 del i.PosAlignList 274 align_dict.build_resnum_list() 275 return (header, sum_dict, align_dict)

276

Source Code for Package Bio.FSSP