Mass-Spectrometry-Based Proteomics Reveals Organ-Specific Expression Patterns To Be Used as Forensic Evidence

作     者：Dammeier, Sascha Nahnsen, Sven Veit, Johannes Wehner, Frank Ueffing, Marius Kohlbacher, Oliver 

作者机构：Univ Hosp Tuebingen Inst Ophthalm Res Med Proteome Ctr D-72074 Tubingen Germany Eberhard Karls Univ Tuebingen Quantitat Biol Ctr D-72076 Tubingen Germany Eberhard Karls Univ Tuebingen Ctr Bioinformat D-72076 Tubingen Germany Eberhard Karls Univ Tuebingen Dept Comp Sci D-72076 Tubingen Germany Eberhard Karls Univ Tuebingen Inst Forens Med D-72074 Tubingen Germany 

出 版 物：《JOURNAL OF PROTEOME RESEARCH》 (蛋白质组研究杂志)

年 卷 期：2016年第15卷第1期

页      面：182-192页

核心收录：

学科分类：0710[理学-生物学] 071010[理学-生物化学与分子生物学] 081704[工学-应用化学] 07[理学] 08[工学] 0817[工学-化学工程与技术] 

基　　金：Deutsche Forschungsgemeinschaft (DFG) [KO-2313/6-1] Deutsche Forschungsgemeinschaft (DFG Institutional Strategy of the University of Tubingen) [ZUK 63] 

主　　题：bottom-up proteomics forensic medicine projectile machine learning algorithm data classification PCA 

摘      要：Standard forensic procedures to examine bullets after an exchange of fire include a mechanical or ballistic reconstruction of the event. While this is routine to identify which projectile hit a subject by DNA analysis of biological material on the surface of the projectile, it is rather difficult to determine which projectile caused the lethal injury often the crucial point with regard to legal proceedings. With respect to fundamental law it is the duty of the public authority to make every endeavor to solve every homicide case. To improve forensic examinations, we present a forensic proteomic method to investigate biological material from a projectile s surface and determine the tissues traversed by it. To obtain a range of relevant samples, different major bovine organs were penetrated with projectiles experimentally. After tryptic on-surface digestion, mass-spectrometry-based proteome analysis, and statistical data analysis, we were able to achieve a cross-validated organ classification accuracy of 99%. Different types of anticipated external variables exhibited no prominent influence on the findings. In addition, shooting experiments were performed to validate the results. Finally, we show that these concepts could be applied to a real case of murder to substantially improve the forensic reconstruction.