[galaxy-user] Gene BED to codon BED not working properly

Hi, When using the gene BED to codon BED tool, I noticed that it is not accurately reporting the codons that make up a gene. For example, some of the codon are missing (particularly ones that span exon-exon junctions. Also, when changing reading frame from one exon to the next, the codons are not read appropriately and the reading frame appears to be decided arbitrarily. Is this a serious known flaw with the tool or am I missing something? Also, is there a version of aaChanges tool that can be used with any genome (not just human genome?). Thank you. Antony