run by run debugging

src/BayesPairing.py

@@ -98,7 +98,8 @@ def parse_sequence(seq,modules,ss,dataset,left_out,sm=0.25, mc=0,k=5000):
                 return_dict[mod] = [this_score,constraints]
 
     return return_dict
-def returner(scores,seq,ss="",m=4,n=4):
+def returner(scores,seq,ss="",m=5,n=5):
+    print("SEQUENCE SCORES:",scores)
     if ss=="":
         maxs = []
         for i in scores.keys():
@@ -109,7 +110,7 @@ def returner(scores,seq,ss="",m=4,n=4):
             returner = []
             final_top = {}
             for z in topitems:
-                if topitems[z]>11:
+                if z[1]>-10:
                     poz = ast.literal_eval(z[0])
                     current_score = z[1]
                     cons = scores[i][1]
@@ -122,7 +123,7 @@ def returner(scores,seq,ss="",m=4,n=4):
                             cons_seq = cons_seq + (cons_dict[nuc])
                         else:
                             cons_seq= cons_seq + ('.')
-                    #print(cons_seq)
+                    print(cons_seq)
                     if "." in cons_seq:
                         #start1 = time.time()
                         nocons_score = testSS.call_rnafold(seq)

src/bayes_to_seqpy.py

@@ -370,7 +370,8 @@ def seq_to_struct(
 
     component_distance = []
     for s in range(len(position_subsets)-1):
-        component_distance.append([150,-1])
+        component_distance.append([5,20])
+    print("OBSERVED COMPONENT DISTANCES")
     for gr in graphs:
         newcols = []
         positions = []
@@ -382,11 +383,12 @@ def seq_to_struct(
 
         for junc in range(len(component_addresses)):
             distance = cols[component_addresses[junc]]-cols[component_addresses[junc]-1]
+            print(distance)
             if distance < component_distance[junc][0]:
                 component_distance[junc][0] = distance
             if distance > component_distance[junc][1]:
                 component_distance[junc][1] = distance
-
+    
     (m, iii, l) = best_struct
    
     print("III")    
@@ -403,7 +405,7 @@ def seq_to_struct(
             if dis[1]<4:
                 dis[1]=5
         if len(re_call) > 0:
-            re_call = re_call + r")([ACGU]{"+str(4)+","+str(dis[1])+"})"
+            re_call = re_call + r")([ACGU]{"+str(dis[0])+","+str(dis[1])+"})"
         if len(i) == 1:
             cons =  [convert[int(k)] for k in list(iii[i[0]])]
             if len(cons) == 1:

src/parse_sequences.py

@@ -26,7 +26,7 @@ def run_fasta(input, modules_to_parse,dataset,ss = "",arguments={}):
     if "sm" in arguments:
         sm=arguments["sm"]
     else:
-        sm=0.2
+        sm=0.9
     if "mc" in arguments:
         mc=arguments["mc"]
     else:
@@ -34,7 +34,7 @@ def run_fasta(input, modules_to_parse,dataset,ss = "",arguments={}):
     if "k" in arguments:
         k=arguments["k"]
     else:
-        k=5000
+        k=50000
     seq = ""
     if "fa" in input:
         prediction_scores = {}
@@ -75,19 +75,20 @@ def run_fasta(input, modules_to_parse,dataset,ss = "",arguments={}):
         else:
             all_maxes=[]
             index = 0
-            while index+200<len(input):
-                print("Running Bayespairing on sequence:",input[index:index+200])
+            while index+50<len(input):
+                print("Running Bayespairing on sequence parts:", index-50,index+50)
                 #maxs=[]
-                maxs = run_BP(input[index:index+200], ss, modules_to_parse, dataset, "NONE",m,n,sm,mc,k)
+                bf = max(0,index-50)
+                maxs = run_BP(input[bf:index+50], ss, modules_to_parse, dataset, "NONE",m,n,sm,mc,k)
                 corrected_maxes = []
                 print(maxs)
                 for ind,cand in enumerate(maxs[0]):
                     if ind%2==1:
-                        corrected_maxes.append([i+index for i in cand])
+                        corrected_maxes.append([i+bf for i in cand])
                     else:
                         corrected_maxes.append(cand)
                 all_maxes.append(corrected_maxes)
-                index=index+200
+                index=index+50
             maxs = run_BP(input[index:], ss, modules_to_parse, dataset, "NONE",m,n,sm,mc,k)
             corrected_maxes = []
             for ind,cand in enumerate(maxs[0]):
@@ -149,5 +150,5 @@ if __name__ == "__main__":
 
     #run_fasta(seq,range(len(graphs)),dataset,ss,arguments)
     #run_fasta(seq,[0,1],dataset,ss,arguments)
-    run_fasta(seq,[2],dataset,ss,arguments)
+    run_fasta(seq,[7],dataset,ss,arguments)