Mercurial > repos > davidmurphy > codonlogo

comparison weblogolib/colorscheme.py @ 0:c55bdc2fb9fa

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author davidmurphy
date Thu, 27 Oct 2011 12:09:09 -0400
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:c55bdc2fb9fa
1
2 # Copyright (c) 2003-2005 The Regents of the University of California.
3 # Copyright (c) 2005 Gavin E. Crooks
4
5 # This software is distributed under the MIT Open Source License.
6 # <http://www.opensource.org/licenses/mit-license.html>
7 #
8 # Permission is hereby granted, free of charge, to any person obtaining a
9 # copy of this software and associated documentation files (the "Software"),
10 # to deal in the Software without restriction, including without limitation
11 # the rights to use, copy, modify, merge, publish, distribute, sublicense,
12 # and/or sell copies of the Software, and to permit persons to whom the
13 # Software is furnished to do so, subject to the following conditions:
14 #
15 # The above copyright notice and this permission notice shall be included
16 # in all copies or substantial portions of the Software.
17 #
18 # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
19 # IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
20 # FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
21 # AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
22 # LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
23 # OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
24 # THE SOFTWARE.
25
26 """ Popular color codings for nucleic and amino acids.
27
28 Classes:
29 ColorScheme -- A color scheme
30 ColorGroup
31
32
33 Generic
34 monochrome
35
36 Nucleotides
37 nucleotide
38 base pairing
39
40 Amino Acid
41 hydrophobicity
42 chemistry
43 charge
44 taylor
45
46 Status : Beta - Needs documentation.
47
48 """
49 # Good online references include bioruby and the JalView alignment editor.
50 # Clamp, M., Cuff, J., Searle, S. M. and Barton, G. J. (2004),
51 # "The Jalview Java Alignment Editor," Bioinformatics, 12, 426-7
52 # http://www.jalview.org
53
54 import sys
55
56 from corebio import seq
57 from color import Color
58 codon_alphabetU=['AAA', 'AAU', 'AAC', 'AAG', 'AUA', 'AUU', 'AUC', 'AUG', 'ACA', 'ACU', 'ACC', 'ACG', 'AGA', 'AGU', 'AGC', 'AGG', 'UAA', 'UAU', 'UAC', 'UAG', 'UUA', 'UUU', 'UUC', 'UUG', 'UCA', 'UCU', 'UCC', 'UCG', 'UGA', 'UGU', 'UGC', 'UGG', 'CAA', 'CAU', 'CAC', 'CAG', 'CUA', 'CUU', 'CUC', 'CUG', 'CCA', 'CCU', 'CCC', 'CCG', 'CGA', 'CGU', 'CGC', 'CGG', 'GAA', 'GAU', 'GAC', 'GAG', 'GUA', 'GUU', 'GUC', 'GUG', 'GCA', 'GCU', 'GCC', 'GCG', 'GGA', 'GGU', 'GGC', 'GGG']
59 codon_alphabetT=['AAA', 'AAT', 'AAC', 'AAG', 'ATA', 'ATT', 'ATC', 'ATG', 'ACA', 'ACT', 'ACC', 'ACG', 'AGA', 'AGT', 'AGC', 'AGG', 'TAA', 'TAT', 'TAC', 'TAG', 'TTA', 'TTT', 'TTC', 'TTG', 'TCA', 'TCT', 'TCC', 'TCG', 'TGA', 'TGT', 'TGC', 'TGG', 'CAA', 'CAT', 'CAC', 'CAG', 'CTA', 'CTT', 'CTC', 'CTG', 'CCA', 'CCT', 'CCC', 'CCG', 'CGA', 'CGT', 'CGC', 'CGG', 'GAA', 'GAT', 'GAC', 'GAG', 'GTA', 'GTT', 'GTC', 'GTG', 'GCA', 'GCT', 'GCC', 'GCG', 'GGA', 'GGT', 'GGC', 'GGG']
60
61 class ColorScheme(object):
62 """ A coloring of an alphabet.
63
64 title : string -- A human readable description
65 defualt_color : Color --
66 groups : list of color groups
67 alphabet : string -- The set of colored symbols
68 color -- A map between a symbol and a Coloring
69
70
71 """
72
73 def __init__(self,
74 groups = [],
75 title = "",
76 description = "",
77 default_color = "black",
78 alphabet = seq.generic_alphabet) :
79 """ """
80 self.title= title
81 self.description = description
82 self.default_color = Color.from_string(default_color)
83 self.groups = groups
84 self.alphabet = alphabet
85 #print >> sys.stderr, groups
86 altype="codons"
87 #print >> sys.stderr,altype
88 #if(alphabet==codon_alphabet):
89 #print >> sys.stderr,"haleyulia it works"
90
91 color = {}
92 #print >> sys.stderr, groups
93 if(alphabet!=codon_alphabetT and alphabet!=codon_alphabetU):
94 for cg in groups :
95 #print >> sys.stderr, cg
96 for s in cg.symbols :
97 color[s] = cg.color
98 #print >> sys.stderr, s
99 #print >> sys.stderr, cg
100 if s not in alphabet :
101 raise KeyError("Colored symbol does not exist in alphabet.")
102 else:
103 for cg in groups :
104 #print >> sys.stderr, cg
105 color[cg.symbols] = cg.color
106 #print >> sys.stderr, cg.symbols
107 self._color = color
108
109 def color(self, symbol) :
110 if symbol in self._color :
111 return self._color[symbol]
112 return self.default_color
113
114 class ColorGroup(object) :
115 """Associate a group of symbols with a color"""
116 def __init__(self, symbols, color, description=None) :
117 self.symbols = symbols
118 self.color = Color.from_string(color)
119 self.description = description
120
121
122
123 monochrome = ColorScheme([]) # This list intentionally left blank
124
125 # From makelogo
126 nucleotide = ColorScheme([
127 ColorGroup("G", "orange"),
128 ColorGroup("TU", "red"),
129 ColorGroup("C", "blue"),
130 ColorGroup("A", "green")
131 ])
132
133 base_pairing = ColorScheme([
134 ColorGroup("TAU", "darkorange", "Weak (2 Watson-Crick hydrogen bonds)"),
135 ColorGroup("GC", "blue", "Strong (3 Watson-Crick hydrogen bonds)")],
136 )
137
138
139 hydrophobicity = ColorScheme([
140 ColorGroup( "RKDENQ", "black", "hydrophobic"),
141 ColorGroup( "SGHTAP", "green", "neutral" ),
142 ColorGroup( "YVMCLFIW", "blue", "hydrophilic") ],
143 alphabet = seq.unambiguous_protein_alphabet
144 )
145
146 # from makelogo
147 chemistry = ColorScheme([
148 ColorGroup( "GSTYC", "green", "polar"),
149 ColorGroup( "NQ", "purple", "neutral"),
150 ColorGroup( "KRH", "blue", "basic"),
151 ColorGroup( "DE", "red", "acidic"),
152 ColorGroup("PAWFLIMV", "black", "hydrophobic") ],
153 alphabet = seq.unambiguous_protein_alphabet
154 )
155
156
157 codonsU = ColorScheme([
158 ColorGroup( 'CAT', '#00FFFF'),
159 ColorGroup( 'CAU', '#00FFFF'),
160 ColorGroup( 'CAC', '#00FFFF'),
161
162 ColorGroup( 'AAA', '#00FFFF'),
163 ColorGroup( 'AAG', '#00FFFF'),
164
165 ColorGroup( 'CGT', '#00FFFF'),
166 ColorGroup( 'CGU', '#00FFFF'),
167 ColorGroup( 'CGC', '#00FFFF'),
168 ColorGroup( 'CGA', '#00FFFF'),
169 ColorGroup( 'CGG', '#00FFFF'),
170 ColorGroup( 'AGA', '#00FFFF'),
171 ColorGroup( 'AGG', '#00FFFF'),
172
173
174 ColorGroup( 'GAT', '#FF0000'),
175 ColorGroup( 'GAU', '#FF0000'),
176 ColorGroup( 'GAC', '#FF0000'),
177
178 ColorGroup( 'GAA', '#FF0000'),
179 ColorGroup( 'GAG', '#FF0000'),
180
181
182 ColorGroup( 'TCT', '#00FF00'),
183 ColorGroup( 'UCU', '#00FF00'),
184 ColorGroup( 'TCC', '#00FF00'),
185 ColorGroup( 'UCC', '#00FF00'),
186 ColorGroup( 'TCA', '#00FF00'),
187 ColorGroup( 'UCA', '#00FF00'),
188 ColorGroup( 'TCG', '#00FF00'),
189 ColorGroup( 'UCG', '#00FF00'),
190 ColorGroup( 'AGT', '#00FF00'),
191 ColorGroup( 'AGU', '#00FF00'),
192 ColorGroup( 'AGC', '#00FF00'),
193
194 ColorGroup( 'ACT', '#00FF00'),
195 ColorGroup( 'ACU', '#00FF00'),
196 ColorGroup( 'ACC', '#00FF00'),
197 ColorGroup( 'ACA', '#00FF00'),
198 ColorGroup( 'ACG', '#00FF00'),
199
200 ColorGroup( 'CAA', '#00FF00'),
201 ColorGroup( 'CAG', '#00FF00'),
202
203 ColorGroup( 'AAT', '#00FF00'),
204 ColorGroup( 'AAU', '#00FF00'),
205 ColorGroup( 'AAC', '#00FF00'),
206
207
208 ColorGroup( 'GCT', '#5555FF'),
209 ColorGroup( 'GCU', '#5555FF'),
210 ColorGroup( 'GCC', '#5555FF'),
211 ColorGroup( 'GCA', '#5555FF'),
212 ColorGroup( 'GCG', '#5555FF'),
213
214 ColorGroup( 'GTT', '#5555FF'),
215 ColorGroup( 'GUU', '#5555FF'),
216 ColorGroup( 'GTC', '#5555FF'),
217 ColorGroup( 'GUC', '#5555FF'),
218 ColorGroup( 'GTA', '#5555FF'),
219 ColorGroup( 'GUA', '#5555FF'),
220 ColorGroup( 'GTG', '#5555FF'),
221 ColorGroup( 'GUG', '#5555FF'),
222
223 ColorGroup( 'CTT', '#5555FF'),
224 ColorGroup( 'CUU', '#5555FF'),
225 ColorGroup( 'CTC', '#5555FF'),
226 ColorGroup( 'CUC', '#5555FF'),
227 ColorGroup( 'CTA', '#5555FF'),
228 ColorGroup( 'CUA', '#5555FF'),
229 ColorGroup( 'CTG', '#5555FF'),
230 ColorGroup( 'CUG', '#5555FF'),
231 ColorGroup( 'TTA', '#5555FF'),
232 ColorGroup( 'UUA', '#5555FF'),
233 ColorGroup( 'TTG', '#5555FF'),
234 ColorGroup( 'UUG', '#5555FF'),
235
236 ColorGroup( 'ATT', '#5555FF'),
237 ColorGroup( 'AUU', '#5555FF'),
238 ColorGroup( 'ATC', '#5555FF'),
239 ColorGroup( 'AUC', '#5555FF'),
240 ColorGroup( 'ATA', '#5555FF'),
241 ColorGroup( 'AUA', '#5555FF'),
242
243 ColorGroup( 'ATG', '#5555FF'),
244 ColorGroup( 'AUG', '#5555FF'),
245
246
247 ColorGroup( 'TTT', '#FF00FF'),
248 ColorGroup( 'UUU', '#FF00FF'),
249 ColorGroup( 'TTC', '#FF00FF'),
250 ColorGroup( 'UUC', '#FF00FF'),
251
252 ColorGroup( 'TAT', '#FF00FF'),
253 ColorGroup( 'UAU', '#FF00FF'),
254 ColorGroup( 'TAC', '#FF00FF'),
255 ColorGroup( 'UAC', '#FF00FF'),
256
257 ColorGroup( 'TGG', '#FF00FF'),
258 ColorGroup( 'UGG', '#FF00FF'),
259
260
261 ColorGroup( 'GGT', '#996600'),
262 ColorGroup( 'GGU', '#996600'),
263 ColorGroup( 'GGC', '#996600'),
264 ColorGroup( 'GGA', '#996600'),
265 ColorGroup( 'GGG', '#996600'),
266
267 ColorGroup( 'CCT', '#996600'),
268 ColorGroup( 'CCU', '#996600'),
269 ColorGroup( 'CCC', '#996600'),
270 ColorGroup( 'CCA', '#996600'),
271 ColorGroup( 'CCG', '#996600'),
272
273
274 ColorGroup( 'TGT', '#FFFF00'),
275 ColorGroup( 'UGU', '#FFFF00'),
276 ColorGroup( 'TGC', '#FFFF00'),
277 ColorGroup( 'UGC', '#FFFF00'),
278
279 ColorGroup( 'TAA', '#000000'),
280 ColorGroup( 'UAA', '#000000'),
281 ColorGroup( 'TAG', '#000000'),
282 ColorGroup( 'UAG', '#000000'),
283 ColorGroup( 'TGA', '#000000'),
284 ColorGroup( 'UGA', '#000000')],
285 alphabet = codon_alphabetU
286 )
287 codonsT = ColorScheme([
288 ColorGroup( 'CAT', '#00FFFF'),
289 ColorGroup( 'CAU', '#00FFFF'),
290 ColorGroup( 'CAC', '#00FFFF'),
291
292 ColorGroup( 'AAA', '#00FFFF'),
293 ColorGroup( 'AAG', '#00FFFF'),
294
295 ColorGroup( 'CGT', '#00FFFF'),
296 ColorGroup( 'CGU', '#00FFFF'),
297 ColorGroup( 'CGC', '#00FFFF'),
298 ColorGroup( 'CGA', '#00FFFF'),
299 ColorGroup( 'CGG', '#00FFFF'),
300 ColorGroup( 'AGA', '#00FFFF'),
301 ColorGroup( 'AGG', '#00FFFF'),
302
303
304 ColorGroup( 'GAT', '#FF0000'),
305 ColorGroup( 'GAU', '#FF0000'),
306 ColorGroup( 'GAC', '#FF0000'),
307
308 ColorGroup( 'GAA', '#FF0000'),
309 ColorGroup( 'GAG', '#FF0000'),
310
311
312 ColorGroup( 'TCT', '#00FF00'),
313 ColorGroup( 'UCU', '#00FF00'),
314 ColorGroup( 'TCC', '#00FF00'),
315 ColorGroup( 'UCC', '#00FF00'),
316 ColorGroup( 'TCA', '#00FF00'),
317 ColorGroup( 'UCA', '#00FF00'),
318 ColorGroup( 'TCG', '#00FF00'),
319 ColorGroup( 'UCG', '#00FF00'),
320 ColorGroup( 'AGT', '#00FF00'),
321 ColorGroup( 'AGU', '#00FF00'),
322 ColorGroup( 'AGC', '#00FF00'),
323
324 ColorGroup( 'ACT', '#00FF00'),
325 ColorGroup( 'ACU', '#00FF00'),
326 ColorGroup( 'ACC', '#00FF00'),
327 ColorGroup( 'ACA', '#00FF00'),
328 ColorGroup( 'ACG', '#00FF00'),
329
330 ColorGroup( 'CAA', '#00FF00'),
331 ColorGroup( 'CAG', '#00FF00'),
332
333 ColorGroup( 'AAT', '#00FF00'),
334 ColorGroup( 'AAU', '#00FF00'),
335 ColorGroup( 'AAC', '#00FF00'),
336
337
338 ColorGroup( 'GCT', '#5555FF'),
339 ColorGroup( 'GCU', '#5555FF'),
340 ColorGroup( 'GCC', '#5555FF'),
341 ColorGroup( 'GCA', '#5555FF'),
342 ColorGroup( 'GCG', '#5555FF'),
343
344 ColorGroup( 'GTT', '#5555FF'),
345 ColorGroup( 'GUU', '#5555FF'),
346 ColorGroup( 'GTC', '#5555FF'),
347 ColorGroup( 'GUC', '#5555FF'),
348 ColorGroup( 'GTA', '#5555FF'),
349 ColorGroup( 'GUA', '#5555FF'),
350 ColorGroup( 'GTG', '#5555FF'),
351 ColorGroup( 'GUG', '#5555FF'),
352
353 ColorGroup( 'CTT', '#5555FF'),
354 ColorGroup( 'CUU', '#5555FF'),
355 ColorGroup( 'CTC', '#5555FF'),
356 ColorGroup( 'CUC', '#5555FF'),
357 ColorGroup( 'CTA', '#5555FF'),
358 ColorGroup( 'CUA', '#5555FF'),
359 ColorGroup( 'CTG', '#5555FF'),
360 ColorGroup( 'CUG', '#5555FF'),
361 ColorGroup( 'TTA', '#5555FF'),
362 ColorGroup( 'UUA', '#5555FF'),
363 ColorGroup( 'TTG', '#5555FF'),
364 ColorGroup( 'UUG', '#5555FF'),
365
366 ColorGroup( 'ATT', '#5555FF'),
367 ColorGroup( 'AUU', '#5555FF'),
368 ColorGroup( 'ATC', '#5555FF'),
369 ColorGroup( 'AUC', '#5555FF'),
370 ColorGroup( 'ATA', '#5555FF'),
371 ColorGroup( 'AUA', '#5555FF'),
372
373 ColorGroup( 'ATG', '#5555FF'),
374 ColorGroup( 'AUG', '#5555FF'),
375
376
377 ColorGroup( 'TTT', '#FF00FF'),
378 ColorGroup( 'UUU', '#FF00FF'),
379 ColorGroup( 'TTC', '#FF00FF'),
380 ColorGroup( 'UUC', '#FF00FF'),
381
382 ColorGroup( 'TAT', '#FF00FF'),
383 ColorGroup( 'UAU', '#FF00FF'),
384 ColorGroup( 'TAC', '#FF00FF'),
385 ColorGroup( 'UAC', '#FF00FF'),
386
387 ColorGroup( 'TGG', '#FF00FF'),
388 ColorGroup( 'UGG', '#FF00FF'),
389
390
391 ColorGroup( 'GGT', '#996600'),
392 ColorGroup( 'GGU', '#996600'),
393 ColorGroup( 'GGC', '#996600'),
394 ColorGroup( 'GGA', '#996600'),
395 ColorGroup( 'GGG', '#996600'),
396
397 ColorGroup( 'CCT', '#996600'),
398 ColorGroup( 'CCU', '#996600'),
399 ColorGroup( 'CCC', '#996600'),
400 ColorGroup( 'CCA', '#996600'),
401 ColorGroup( 'CCG', '#996600'),
402
403
404 ColorGroup( 'TGT', '#FFFF00'),
405 ColorGroup( 'UGU', '#FFFF00'),
406 ColorGroup( 'TGC', '#FFFF00'),
407 ColorGroup( 'UGC', '#FFFF00'),
408
409 ColorGroup( 'TAA', '#000000'),
410 ColorGroup( 'UAA', '#000000'),
411 ColorGroup( 'TAG', '#000000'),
412 ColorGroup( 'UAG', '#000000'),
413 ColorGroup( 'TGA', '#000000'),
414 ColorGroup( 'UGA', '#000000')],
415 alphabet = codon_alphabetT
416 )
417
418
419
420
421 charge = ColorScheme([
422 ColorGroup("KRH", "blue", "Positive" ),
423 ColorGroup( "DE", "red", "Negative") ],
424 alphabet = seq.unambiguous_protein_alphabet
425 )
426
427
428 taylor = ColorScheme([
429 ColorGroup( 'A', '#CCFF00' ),
430 ColorGroup( 'C', '#FFFF00' ),
431 ColorGroup( 'D', '#FF0000'),
432 ColorGroup( 'E', '#FF0066' ),
433 ColorGroup( 'F', '#00FF66'),
434 ColorGroup( 'G', '#FF9900'),
435 ColorGroup( 'H', '#0066FF'),
436 ColorGroup( 'I', '#66FF00'),
437 ColorGroup( 'K', '#6600FF'),
438 ColorGroup( 'L', '#33FF00'),
439 ColorGroup( 'M', '#00FF00'),
440 ColorGroup( 'N', '#CC00FF'),
441 ColorGroup( 'P', '#FFCC00'),
442 ColorGroup( 'Q', '#FF00CC'),
443 ColorGroup( 'R', '#0000FF'),
444 ColorGroup( 'S', '#FF3300'),
445 ColorGroup( 'T', '#FF6600'),
446 ColorGroup( 'V', '#99FF00'),
447 ColorGroup( 'W', '#00CCFF'),
448 ColorGroup( 'Y', '#00FFCC')],
449 title = "Taylor",
450 description = "W. Taylor, Protein Engineering, Vol 10 , 743-746 (1997)",
451 alphabet = seq.unambiguous_protein_alphabet
452 )
453
454
455