Our analysis of structures of the proline biosynthetic enzyme PYCR1 deposited by another group in 2006 suggested that the NAD(P)H cofactor had been erroneously modeled into noise density, placing the ligand over 25 Angstrom from the canonical binding site for Rossmann fold enzymes. Because of our longstanding interest in proline metabolism, I felt compelled to set the record straight.  Therefore, we generated a clone, purified the protein, and determined several structures with different ligands bound. Sure enough, the ligands in the 2006 structures were incorrect.  The supplement to our paper includes a thorough analysis of the electron density of the incorrect structures (EDS maps, PDBREDO, and omit maps).  We also did kinetics and analytical ultracentrifugation studies.  In short, we did a fairly comprehensive biochemical and biophysical study of the enzyme.  Thus, pointing out someone else’s error was just one part of a larger paper.