AMELY deletion is not detected in systematically sampled reference populations: A Reply to Štamfelj

Biomolecular sex estimation promises to fill a major gap in the bioarchaeological record by providing estimates of biological sex for skeletal remains with degraded or ambiguous osteological sex-specific markers. Genomic and proteomic sex estimation, like all analytical methods, have limitations and require frameworks to address the problems of low signal samples and the inevitable conflicting results when other methods are used. Proteomic sex estimation is based on the detection of sex-chromosome specific amelogenin protein fragments in enamel using mass spectrometry. Enamel from male individuals contains amelogenin fragments from both the X-and Y-chromosome versions of amelogenin, and enamel from female individuals contains fragments from only the X-chromosome protein. The method is sensitive, robust, quantifiable and reproducible. Researchers have developed, and continue to develop, frameworks to address theoretical problems associated with low levels of detection and conflicting sex estimates that will inevitably occur when multiple methods are used on a sufficiently large dataset. Štamfelj reminds readers that structural variants of the Y-chromosome that delete the amelogenin gene have been detected in forensics and clinical casework. Since this phenomenon would also account for the absence of the AMELY protein in enamel it should therefore be mentioned as an alternative hypothesis by investigators, along with female sex and low peptide signals in mass spectrometry. In his meta-analysis Štamfelj concludes that this is an intrinsic limitation of biomolecular sex estimation, particularly when examining South Asian populations, and should be incorporated in standard analytical sex estimation frameworks. In this comment, we test this assertion by examining the occurrence of AMELY deletion in the systematically sampled, high coverage, large scale, and well-curated populations of the 1000 Genomes Project and Exome Sequencing Project. When using SNP loci in the open reading frame of AMELY, structural deletion was not detected in either project. Confident probabilities of occurrence with associated intervals cannot be determined from null values. We conclude from this that, for now, AMELY deletion should have no bearing on routine biomolecular sex estimation.