A new study published in Nature by Bayer Pharmaceuticals shows that a big percentage of interesting new results in medical, biology, or med-chem literature are impossible to reproduce, reports In the Pipeline's Derek Lowe.