Using simulated microhaplotype genotyping data to evaluate the value of machine learning algorithms for inferring DNA mixture contributor numbers.

Inferring the number of contributors (NoC) is a crucial step in interpreting DNA mixtures, as it directly affects the accuracy of the likelihood ratio calculation and the assessment of evidence strength. However, obtaining the correct NoC in complex DNA mixtures remains challenging due to the high degree of allele sharing and dropout. This study aimed to analyze the impact of allele sharing and dropout on NoC inference in complex DNA mixtures when using microhaplotypes (MH). The effectiveness and value of highly polymorphic MH for NoC inference in complex DNA mixtures were evaluated through comparing the performance of three NoC inference methods, including maximum allele count (MAC) method, maximum likelihood estimation (MLE) method, and random forest classification (RFC) algorithm. In this study, we selected the top 100 most polymorphic MH from the Southern Han Chinese (CHS) population, and simulated over 40 million complex DNA mixture profiles with the NoC ranging from 2 to 8. These profiles involve unrelated individuals (RM type) and related pairs of individuals, including parent-offspring pairs (PO type), full-sibling pairs (FS type), and second-degree kinship pairs (SE type). Our results indicated that how the number of detected alleles in DNA mixture profiles varied with the markers' polymorphism, kinship's involvement, NoC, and dropout settings. Across different types of DNA mixtures, the MAC and MLE methods performed best in the RM type, followed by SE, FS, and PO types, while RFC models showed the best performance in the PO type, followed by RM, SE, and FS types. The recall of all three methods for NoC inference were decreased as the NoC and dropout levels increased. Furthermore, the MLE method performed better at low NoC, whereas RFC models excelled at high NoC and/or high dropout levels, regardless of the availability of a priori information about related pairs of individuals in DNA mixtures. However, the RFC models which considered the aforementioned priori information and were trained specifically on each type of DNA mixture profiles, outperformed RFC_ALL model that did not consider such information. Finally, we provided recommendations for model building when applying machine learning algorithms to NoC inference.

Calculation of the Paternity Index for the Alleged Father Related to the Child's Mother.

OBJECTIVES: To derive the paternity index (PI) calculation formula of the alleged father (AF) when the AF is a relative (parent/child, siblings, grandparent/grandchild, uncle/nephew, first cousins) of the child's biological mother. METHODS: For the case when the AF is related to the child's biological mother, the existence of the relationship in the numerator and denominator hypothesis of PI was considered. The genotype frequency of the AF was calculated by using the frequency formula in which the mother's genotype was considered, while the random male in the denominator was substituted as another relative of the mother's same rank. The PI calculation formula was derived to eliminate the effect of the relationship between AF and the child's biological mother. RESULTS: When the AF and the biological mother have first, second and tertiary kinship, a more conservative PI was obtained from the PI calculation formula derived in this study compared with the PI calculation method which did not consider kinship. CONCLUSIONS: The calculation method provided in this study can eliminate the effect of the relation of the AF and mother on the PI in incest cases, to obtain more accurate and conservative identification conclusions.

A proof-of-principle study on implementing polymerase chain displacement reaction (PCDR) to improve forensic low-template DNA analysis.

Polymerase chain reaction (PCR) plays an important role in forensic DNA analysis. However, the amplification of low-template DNA (LTDNA) samples usually encounters unsatisfactory results for the limited efficiency of PCR, which would interfere with the subsequent profile interpretation. Polymerase chain displacement reaction (PCDR) is a highly-efficient technique characterized by combining PCR and strand displacement reaction into a single PCDR cycle. This study explored the feasibility of PCDR for improving forensic LTDNA analysis. STR markers commonly used in forensic genetics were subjected to PCDR amplification and capillary electrophoresis detection. The results of singleplex reactions indicated that PCDR surpassed original PCR in efficiency for STR amplification. The average peak height of alleles in PCDR profiles was linearly correlated to the number of outer primers adopted for initiating the strand displacement process. Further, we assessed the multiplexing potential of PCDR by incorporating 17 STRs included in the expanded CODIS core loci and Amelogenin gene into a multiplex PCDR system. For pristine DNA templates ranged from 200 pg to 12.5 pg, the multiplex PCDR system consistently exhibited higher allele peak height as well as less allele dropout compared to the multiplex PCR references. Meanwhile, a significant reduction of stutter ratio was extensively observed in PCDR profiles. We also tested mock casework samples to verify the practical ability of multiplex PCDR for LTDNA detection. With DNA input varying from 48.1 pg to 6.6 pg, the multiplex PCDR system consistently obtained more allelic information than multiplex PCR methods. Our data collectively suggested that it is feasible to apply PCDR in forensic LTDNA analysis.

Evaluation of microfluidics-based droplet PCR combined with multiplex STR system in forensic science.

Because of its excellent monodispersity, high throughput, and low volume, microfluidics-based droplet PCR has become the core technology of digital PCR, next-generation sequencing, and other technology platforms. This study constructed a microfluidic water-in-oil droplet PCR system and amplified a commercially available forensic 22-plex short tandem repeat detection system. We analyzed the sensitivity, concordance, amplification efficiency of the droplet PCR, and influence factors of the above aspects. The droplet PCR showed high concordance with conventional bulk PCR and had high sensitivity as 0.125 ng. Furthermore, we observed the performance of droplet PCR in high-order mixed DNA. As the mixture ratios from 10:1 to 30:1, droplet PCR presented more mixture proportion (Mx) increased loci from 11 (57.89%) to 17 (89.47%). In the mixture ratios 20:1, 25:1, and 30:1, significant Mx differences between droplet PCR and bulk PCR were observed (p < 0.05). The results showed that the droplet PCR could improve the identification of the minor contributor's DNA in a two-person mixture and alleviate the imbalanced amplification problem. This study provides a reference and basis for the wide application of droplet PCR in forensic science.

Forensic genomics research on microhaplotypes.

Microhaplotype loci (microhaplotype, MHs), defined by two or more closely linked single nucleotide polymorphisms, are a type of molecular marker within a short segment of DNA. As emerging forensic genetic markers, MHs have no stutter artefacts and higher polymorphism, and permit the design of smaller amplicons. In order to identify the markers from a genome wide perspective and explore their potential application further, we constructed the most comprehensive MH dataset to date, based on the whole genome sequencing data of 105 Han individuals in Southern China from 1000 Genomes Project. The results showed that there were 9,490,075 MH loci in the range of 350 bp in the human genome, and the distribution density of microhaplotypes suggests gene variation. Polymorphism analysis of MHs from various base spans showed that the polymorphism of MHs could reach or exceed common short tandem repeat sites. In addition, based on their flexible assembly, a scheme to build the public database of microhaplotypes was proposed.