Catch zero-counts in codon frequency reporting

If no codons counted, skip amino acid consensus summary.

Use dummy value of 0.000001 minimum freq for amino acids not observed in
sample, for cases where sample too small to pick up all amino acids at
least once.

porc_cod_usage.py

@@ -312,12 +312,15 @@ def ok_cod(acod):
 codcons = {}
 for cod in all_hist:
   thres50 = sum([all_hist[cod][aa] for aa in all_hist[cod]])/2
-  candidate = [aa for aa in all_hist[cod] if all_hist[cod][aa] >= thres50]
-  if len(candidate) >= 1:
-      # concatenate if there is a 50-50 tie (unlikely except for singleton counts)
-      codcons[cod] = '_'.join(candidate)
+  if thres50 == 0: # no codons counted!
+    codcons[cod] = '-'
   else:
-      codcons[cod] = '?'
+    candidate = [aa for aa in all_hist[cod] if all_hist[cod][aa] >= thres50]
+    if len(candidate) >= 1:
+        # concatenate if there is a 50-50 tie (unlikely except for singleton counts)
+        codcons[cod] = '_'.join(candidate)
+    else:
+        codcons[cod] = '?'
 
 # Summary statistics on codon frequencies in alignments
 with open(args.counts, 'w') as fh:
@@ -359,7 +362,7 @@ def ok_cod(acod):
     # matrix for Weblogo plotting
     if stop_tot[cod] > args.codon_threshold * mean(list(other_tot.values())):
       outlines.append(
-        "\t".join([str(i)] + [str(float(all_hist[cod][aa])/aa_freq_d[aa]) for aa in aa_list]) + "\n")
+        "\t".join([str(i)] + [str(float(all_hist[cod][aa])/(aa_freq_d[aa] + 0.00001)) for aa in aa_list]) + "\n")
       #outlines.append("%s" % i + "\t" + "\t".join([str(float(all_hist[cod][aa])) for aa in aa_list]) + "\n")
     # else append zeroes (column not displayed by Weblogo)
     else: