Augustus Parsing Loci names and gene names from a Gene bank file format

Question

[mwerseb1]

Hi All,

I am following along the Augustus tutorial in Current Protocols in Bioinformatics (https://currentprotocols.onlinelibrary.wiley.com/doi/full/10.1002/cpbi.57). I am working through Protocol 6- Removing Redundant Gene Structures, and have run into an issue on Step 4- creating a loci.lst file. When I run the provided perl code snippet (see below) I get the error message below. I am not familiar with perl and would greatly appreciate if anyone can help figure out how to get this section to work.

Code Snippet:

cat bonafide.gb | perl -ne ’
if ( $_ =~ m/LOCUS\s+(\S+)\s/ ) {
$txLocus = $1;
} elsif ( $_ =~ m/\/gene=\"(\S+)\"/ ) {
$txInGb3{$1} = $txLocus
}
if( eof() ) {
foreach ( keys %txInGb3 ) {
print "$_\t$txInGb3{$_}\n";
}
}’ > loci.lst

where the bonafide.gb file looks like this: I need to pull out the locus name "h2tg000001l_432666-437116" and the gene name "h2tg000001l_t_gene1_mRNA1"

LOCUS       h2tg000001l_432666-437116   4451 bp  DNA
FEATURES             Location/Qualifiers
     source          1..4451
     mRNA            join(1286..1450,1766..1909,2591..3166)
                     /gene="h2tg000001l_t_gene1_mRNA1"
     CDS             join(1286..1450,1766..1909,2591..3166)
                     /gene="h2tg000001l_t_gene1_mRNA1"
BASE COUNT     599 a   460 c  353 g   518 t   2521 n
ORIGIN
        1 nnnnnnnnnn nnnnnnnnnn nnnnnnnnnn nnnnnnnnnn nnnnnnnnnn nnnnnnnnnn
       61 nnnnnnnnnn nnnnnnnnnn nnnnnnnnnn nnnnnnnnnn nnnnnnnnnn nnnnnnnnnn
      121 nnnnnnnnnn nnnnnnnnnn nnnnnnnnnn nnnnnnnnnn nnnnnnnnnn nnnnnnnnnn

Error message:

Unrecognized character \xE2; marked by <-- HERE after <-- HERE near column 1 at -e line 1.
./make_loci_list.sh: line 4: syntax error near unexpected token `('
./make_loci_list.sh: line 4: `if ( $_ =~ m/LOCUS\s+(\S+)\s/ ) {'

Any help in either debugging or any way to get a similar result with bash would be incredibly helpful.

Many Thanks!

Answer 1

[h.mon]

Did you copy / paste the perl code directly from the pdf? The problem is you have one (two, actually) invalid characters in the code snippet you copied, the ’ enclosing the perl code - it should be '.