[maker-devel] [PATCH] Corrupted exon table file when running FGeneSH

[Neil Moore]

I encountered problems with FGeneSH failing on some contigs (small
ones with few if any genes).  Investigating the logs, I found that
fgenesh was complaining about a "Corrupted exon table file"; it turns
out that MAKER had omitted the header line from the exon table.  I
think this happened when there were predictions or evidence for that
contig, but none of them contained introns; I haven't been able to
verify that, though.

The following patch corrected the problem and allowed FGeneSH to run,
but I don't know the code well so perhaps there is a better way.


--- lib/Widget/fgenesh.pm	2013-05-22 12:34:00.000000000 -0400
+++ /home/neil/fgenesh.pm	2013-08-08 14:41:40.585139335 -0400
@@ -562,18 +562,18 @@
 	   push(@xdef, $l);
         }
 
-        return \@xdef if(!@$i_coors);            
-    
-       foreach my $i (@$i_coors){
-	   my $i_b = ($i->[0] - $offset) + ($i_flank-1);
-	   my $i_e = ($i->[1] - $offset) - ($i_flank-1);
-                
-	   next if abs($i_b - $i_e) < 2*$i_flank;
-	   next if abs($i_b - $i_e) < 25;
-	   
-	   my $l  = "$i_b $i_e -1000";
-	   
-	   push(@xdef, $l);
+       if(@$i_coors) {
+	   foreach my $i (@$i_coors){
+	       my $i_b = ($i->[0] - $offset) + ($i_flank-1);
+	       my $i_e = ($i->[1] - $offset) - ($i_flank-1);
+
+	       next if abs($i_b - $i_e) < 2*$i_flank;
+	       next if abs($i_b - $i_e) < 25;
+
+	       my $l  = "$i_b $i_e -1000";
+
+	       push(@xdef, $l);
+	   }
        }
 
        my $num = @xdef;


-- 
Dr Neil Moore, neil at cs.uky.edu, neil at uky.edu, neil at s-z.org