Brazilian forensic casework analysis through MPS applications: Statistical weight-of-evidence and biological nature of criminal samples as an influence factor in quality metrics

• Published in: Forensic science international 2019-10, Vol.303, p.109938-109938, Article 109938
• Main Authors: Avila, E., Cavalheiro, C.P., Felkl, A.B., Graebin, P., Kahmann, A., Alho, C.S.
• Format: Article
• Language: English
• Subjects: Alleles; Blood Chemical Analysis; Brazil; Brazilian Federal Police; Capillary electrophoresis; Condoms; Crime; Data analysis; Databases, Genetic; Deoxyribonucleic acid; DNA; DNA fingerprinting; DNA Fingerprinting - methods; Electrophoresis; Electrophoresis, Capillary; Epithelial Cells; Evaluation; Forensic; Forensic science; Forensic sciences; Gene Frequency; Genetics, Population; Genotyping; HID ion Ampliseq identity panel; High-Throughput Nucleotide Sequencing; Humans; Identification; Inspection; Investigations; Massive Parallel Sequencing; Microsatellite Repeats; Nature of biological evidence; Polymorphism, Single Nucleotide; Quality; Quality metrics; Quantitation; Real-Time Polymerase Chain Reaction; Reflexes; Reliability analysis; Saliva; Samples; Semen; Sequence Analysis, DNA; Sequences; SNP; Statistical analysis; Statistical methods; Statistical tests; Substrates; Trustworthiness; Weight
• ISSN: 0379-0738; 1872-6283
• DOI: 10.1016/j.forsciint.2019.109938

•Real forensic casework samples were successfully evaluated through MPS techniques.•Statistical strength of evidence of SNP panel outperforms traditional commercial STR kits.•Use of regional allele frequency databases shows no statistical difference to national data.•Biological nature of evidence seems to impact quality metrics in MPS applications.•Epithelial samples presented lowest quality results among types of biological evidence. Real forensic casework biological evidence can be found in a myriad of different conditions and presenting very distinct features, including key elements such as degradation levels, the nature of biological evidence, mixture presence, and surface or substrate deposition, among others. Technical protocols employed by forensic DNA analysts must consider such characteristics in order to improve the chances of successfully genotyping these materials. MPS has been used as a very useful tool for forensic sample processing and genetic profile generation. However, it is not completely clear how different features encountered with real forensic samples impact sequencing quality and, consequently, profile accuracy and reliability. In this context, the present study analyzes a set of 47 real forensic casework samples, obtained from semen, saliva, blood and epithelial evidence, as well as reference oral swabs, aiming to evaluate the impact of a sample’s biological nature in profiling success. All DNA extracts from samples were standardized according to sample conditions, as assessed by traditional forensic profiling methods (real-time PCR quantitation and capillary electrophoresis-coupled STR fragment analysis). Samples were separated into groups according to their biological nature, and the resultant sequencing quality was evaluated through a series of well-established statistical tests, applied specifically to six different MPS quality metrics. The results showed that certain groups of samples, especially epithelial and (to a lesser extent) saliva samples, exhibited significantly lower quality in terms of some of the evaluated metrics. A number of reasons for such unexpected behavior are discussed. In addition, a series of calculations was performed to assess the weight of genetic evidence in Brazilian samples, and reflexes in data analysis and national allele frequency database construction are discussed. Overall, the results indicate that a unified national allele frequency database can be used nationwide. Besides this, MPS genetic p