ASR_added

prob.py

@@ -203,9 +203,19 @@ class Distrib():
                 maxprob = self[sym]
         return maxsym
 
+    def getProb(self):
+        """ Return a list of (sym, prob) tuples, in order of their alphabet"""
+        return [(sym, self.prob(sym)) for sym in self.alpha]
+
+    def getBits(self):
+        """ Return a list of (sym, bits) tuples, in order of their alphabet"""
+        H = sum([-f * (math.log2(f) if f > 0 else 999) for f in self.prob()])
+        I = math.log2(len(self.alpha)) - H
+        return [(s, self.prob(s) * I) for s in self.alpha]
+
     def getsort(self):
         """ Return the list of symbols, in order of their probability. """
-        symlist = [sym for (sym, _) in self.getProbsort()]
+        symlist = [s for (s, _) in self.getProbsort()]
         return symlist
 
     def getProbsort(self):
@@ -1089,3 +1099,18 @@ def lgamma(x):
         y += 1
         ser += (cof[j] / y)
     return (-tmp + math.log(2.5066282746310005 * ser / x))
+
+
+import sequence as seq
+if __name__ == '__main__':
+    myseqs = [seq.Sequence('TCCTAGCCCC'),
+              seq.Sequence('GCCGCCCCCA'),
+              seq.Sequence('ATCCGCCCGG'),
+              seq.Sequence('CCCCCGCCTT')]
+    mymc = MarkovChain(seq.DNA_Alphabet)
+    for myseq in myseqs:
+        print(myseq, len(myseq))
+        mymc.observe(myseq)
+    for t in mymc.transit:
+        print(t, mymc.transit[t])
+

sequence.py

@@ -1344,6 +1344,61 @@ class PWM(object):
                 maxindex = i
         return (maxscore, maxindex)
 
+def readPWMs(filename, format='MEME'):
+    fh = open(filename, 'rt')
+    VERSION = None
+    ALPHABET = None
+    STRANDS = None
+    BACKGROUND = None
+    MOTIF = None
+    URL = None
+    FOREGROUND = []
+    NSITES = 1
+    EXPECT = None
+    COLLECTION = {}
+    data = fh.read()
+    lines = data.splitlines()
+    for line in lines:
+        myline = line.strip()
+        words = myline.split();
+        if EXPECT == None:
+            try:
+                if myline.startswith('MEME') and len(words) > 1:
+                    VERSION = words[len(words) - 1]
+                elif myline.startswith('ALPHABET') and len(words) > 1:
+                    ALPHABET = Alphabet(words[len(words) - 1])
+                elif myline.startswith('MOTIF') and len(words) > 1:
+                    if MOTIF != None: # we have one motif that needs to be stored first
+                        COLLECTION[MOTIF] = (FOREGROUND, BACKGROUND)
+                    FOREGROUND = []
+                    MOTIF = words[1]
+                elif myline.startswith('URL') and len(words) > 1:
+                    URL = words[1]
+                elif len(words) == 0:
+                    EXPECT = None
+                elif not myline.startswith('#'):
+                    EXPECT = myline
+                    if EXPECT.startswith('letter-probability matrix'):
+                        match = re.compile(r""".*nsites=\s*(?P<name>[0-9]*?)\s.*""", re.VERBOSE).match(EXPECT)
+                        NSITES = int(match.group('name'))
+            except:
+                print('Error in format: ' + line)
+                return None
+        elif len(words) == 0:
+            EXPECT = None
+        elif EXPECT.startswith('Background'):
+            BACKGROUND = Distrib(ALPHABET)
+            for z in zip(words[::2], words[1::2]):
+                BACKGROUND.observe(z[0], float(z[1]))
+        elif EXPECT.startswith('letter-probability matrix'):
+            d = Distrib(ALPHABET)
+            for z in zip(ALPHABET.symbols, words):
+                d.observe(z[0], float(z[1]) * NSITES)
+            FOREGROUND.append(d)
+    # save last motif here
+    COLLECTION[MOTIF] = (FOREGROUND, BACKGROUND)
+    return COLLECTION
+
 # Web Service Functions -------------------
 
 def getSequence(id, database = 'uniprotkb', start=None, end=None):
@@ -1438,7 +1493,7 @@ def runBLAST(sequence, program='blastp', database='uniprotkb', exp='1e-1'):
             ids.append(id.split(':')[1])
     return ids
 
-if __name__ == '__main__':
+if __name__ == '__main__1':
     aln = readClustalFile('/Users/mikael/simhome/ASR/gappy.aln', Protein_Alphabet)
     x, g, i = aln.outliers()
     for s in range(len(aln)):
@@ -1449,3 +1504,10 @@ if __name__ == '__main__':
         if idx >= 0:
             print('\t', aln[s].sequence[idx:])
     print(('Read', len(aln), 'sequences'))
+
+if __name__ == '__main__':
+    motifs = readPWMs('/Users/mikael/meme-5.4.1/motif_databases/PROKARYOTE/collectf.meme')
+    for name in motifs:
+        print(name)
+        for fg in motifs[name][0]:
+            print('\t', fg)
\ No newline at end of file