Number of CDS changes when I re-annotate a rotated chromosome

[jhammery]

Hallo,

I used Prokka via Galaxy Europe to annotate de novo bacterial genome assemblies. I loaded the genbank-files into Clone Manager in order to rotate the molecule so that the dnaA gene (according to the initial Prokka annotation) is the first open reading frame. Then I exported the molecule as fasta and re-annotated the genome using Prokka via Galaxy Europe. Surprisingly, this resulted in a decreased number of CDS (the genome size did not change) for both assemblies.

1st assembly before rotation to dnaA organism: Genus species strain contigs: 1 bases: 9719866 CDS: 7060 gene: 7144 rRNA: 3 repeat_region: 2 tRNA: 80 tmRNA: 1

1st assembly after rotation to dnaA organism: Genus species strain contigs: 1 bases: 9719866 CDS: 7040 gene: 7124 rRNA: 3 repeat_region: 2 tRNA: 80 tmRNA: 1

2nd assembly before rotation to dnaA organism: Genus species strain contigs: 1 bases: 7031387 CDS: 5560 gene: 5639 rRNA: 8 repeat_region: 1 tRNA: 71

2nd assembly after rotation to dnaA* organism: Genus species strain contigs: 1 bases: 7031387 CDS: 5557 gene: 5636 rRNA: 8 repeat_region: 1 tRNA: 71

I have also noticed that in one case, the dnaA gene was now 18 bp shorter. As the DNA sequence is the same (only rotated) I expected that the annotation would be identical and reproducible.

Why is this happening and how can I avoid it in the future? Thank you in advance for your help.

Best regards,

Jonathan