Jaroslav Brůžek1,2, Frédéric Santos1, Bruno Dutailly1, Pascal Murail1, Eugenia Cunha3. 1. Laboratoire PACEA - De la Préhistoire à l'Actuel: Culture, Environnement et Anthropologie, UMR 5199, CNRS, Université de Bordeaux, CS 50023, Pessac, 33615, France. 2. Department of Anthropology and Human Genetics, Faculty of Science, Charles University, Prague 2, 12000, Czech Republic. 3. Centre for Functional Ecology, Laboratory of Forensic Anthropology, Life Sciences Department, University of Coimbra, Coimbra, 3000-456, Portugal.
Abstract
OBJECTIVES: A new tool for skeletal sex estimation based on measurements of the human os coxae is presented using skeletons from a metapopulation of identified adult individuals from twelve independent population samples. For reliable sex estimation, a posterior probability greater than 0.95 was considered to be the classification threshold: below this value, estimates are considered indeterminate. By providing free software, we aim to develop an even more disseminated method for sex estimation. MATERIALS AND METHODS: Ten metric variables collected from 2,040 ossa coxa of adult subjects of known sex were recorded between 1986 and 2002 (reference sample). To test both the validity and reliability, a target sample consisting of two series of adult ossa coxa of known sex (n = 623) was used. The DSP2 software (Diagnose Sexuelle Probabiliste v2) is based on Linear Discriminant Analysis, and the posterior probabilities are calculated using an R script. RESULTS: For the reference sample, any combination of four dimensions provides a correct sex estimate in at least 99% of cases. The percentage of individuals for whom sex can be estimated depends on the number of dimensions; for all ten variables it is higher than 90%. Those results are confirmed in the target sample. DISCUSSION: Our posterior probability threshold of 0.95 for sex estimate corresponds to the traditional sectioning point used in osteological studies. DSP2 software is replacing the former version that should not be used anymore. DSP2 is a robust and reliable technique for sexing adult os coxae, and is also user friendly.
OBJECTIVES: A new tool for skeletal sex estimation based on measurements of the human os coxae is presented using skeletons from a metapopulation of identified adult individuals from twelve independent population samples. For reliable sex estimation, a posterior probability greater than 0.95 was considered to be the classification threshold: below this value, estimates are considered indeterminate. By providing free software, we aim to develop an even more disseminated method for sex estimation. MATERIALS AND METHODS: Ten metric variables collected from 2,040 ossa coxa of adult subjects of known sex were recorded between 1986 and 2002 (reference sample). To test both the validity and reliability, a target sample consisting of two series of adult ossa coxa of known sex (n = 623) was used. The DSP2 software (Diagnose Sexuelle Probabiliste v2) is based on Linear Discriminant Analysis, and the posterior probabilities are calculated using an R script. RESULTS: For the reference sample, any combination of four dimensions provides a correct sex estimate in at least 99% of cases. The percentage of individuals for whom sex can be estimated depends on the number of dimensions; for all ten variables it is higher than 90%. Those results are confirmed in the target sample. DISCUSSION: Our posterior probability threshold of 0.95 for sex estimate corresponds to the traditional sectioning point used in osteological studies. DSP2 software is replacing the former version that should not be used anymore. DSP2 is a robust and reliable technique for sexing adult os coxae, and is also user friendly.
Authors: Jana Velemínská; Nikola Fleischmannová; Barbora Suchá; Jan Dupej; Šárka Bejdová; Anežka Kotěrová; Jaroslav Brůžek Journal: Int J Legal Med Date: 2021-03-01 Impact factor: 2.686
Authors: Maïté Rivollat; Aline Thomas; Emmanuel Ghesquière; Adam Benjamin Rohrlach; Ellen Späth; Marie-Hélène Pemonge; Wolfgang Haak; Philippe Chambon; Marie-France Deguilloux Journal: Proc Natl Acad Sci U S A Date: 2022-04-21 Impact factor: 12.779
Authors: Meucci M; Costedoat C; Verna E; Adam F; Signoli M; Drancourt M; Beye M; Aboudharam G; Barbieri R Journal: Curr Res Microb Sci Date: 2021-12-04
Authors: Rebeka Rmoutilová; Pierre Guyomarc'h; Petr Velemínský; Alena Šefčáková; Mathilde Samsel; Frédéric Santos; Bruno Maureille; Jaroslav Brůžek Journal: PLoS One Date: 2018-08-30 Impact factor: 3.240