A former research fellow at Harvard School of Dental Medicine falsified numbers in slides presented at a laboratory meeting and in research summaries submitted to conferences, according to a finding by the federal Office of Research Integrity.

The scientist, Martin Biosse-Duplan, switched numbers to show differences in bone density between genetically-modified mice and a comparison group of mice, according to the notice, published on Friday. The finding was based on an investigation by Harvard Medical School and Harvard School of Dental Medicine.