Antoinette A. Westen

Low template STR typing: Effect of replicate number and consensus method on genotyping reliability and DNA database search results

To analyze DNA samples with very low DNA concentrations, various methods have been developed that sensitize short tandem repeat (STR) typing. Sensitized DNA typing is accompanied by stochastic amplification effects, such as allele drop-outs and drop-ins. Therefore low template (LT) DNA profiles are interpreted with care. One can either try to infer the genotype by a consensus method that uses alleles confirmed in replicate analyses, or one can use a statistical model to evaluate the strength of the evidence in a direct comparison with a known DNA profile. In this study we focused on the first strategy and we show that the procedure by which the consensus profile is assembled will affect genotyping reliability. In order to gain insight in the roles of replicate number and requested level of reproducibility, we generated six independent amplifications of samples of known donors. The LT methods included both increased cycling and enhanced capillary electrophoresis (CE) injection [1]. Consensus profiles were assembled from two to six of the replications using four methods: composite (include all alleles), n-1 (include alleles detected in all but one replicate), n/2 (include alleles detected in at least half of the replicates) and 2× (include alleles detected twice). We compared the consensus DNA profiles with the DNA profile of the known donor, studied the stochastic amplification effects and examined the effect of the consensus procedure on DNA database search results. From all these analyses we conclude that the accuracy of LT DNA typing and the efficiency of database searching improve when the number of replicates is increased and the consensus method is n/2. The most functional number of replicates within this n/2 method is four (although a replicate number of three suffices for samples showing >25% of the alleles in standard STR typing). This approach was also the optimal strategy for the analysis of 2-person mixtures, although modified search strategies may be needed to retrieve the minor component in database searches. From the database searches follows the recommendation to specifically mark LT DNA profiles when entering them into the DNA database.

Higher capillary electrophoresis injection settings as an efficient approach to increase the sensitivity of STR typing.

Abstract:  Evidentiary traces may contain low quantities of DNA, and regularly incomplete short tandem repeat (STR) profiles are obtained. In this study, higher capillary electrophoresis injection settings were used to efficiently improve incomplete STR profiles generated from low‐level DNA samples under standard polymerase chain reaction (PCR) conditions. The method involves capillary electrophoresis with higher injection voltage and extended injection time. STR peak heights increased six‐fold. Inherent to the analysis of low‐level DNA samples, we observed stochastic amplification artifacts, mainly in the form of allele dropout and heterozygous peak imbalance. Increased stutter ratios and allele drop‐in were rarely seen. Upon STR typing of 10:1 admixed samples, the profile of the major component did not become overloaded when using higher injection settings as was observed upon elevated cycling. Thereby an improved profile of the minor component was obtained. For low‐level DNA casework samples, we adhere to independent replication of the PCR amplification and boosted capillary electrophoresis.

Tri-allelic SNP markers enable analysis of mixed and degraded DNA samples

For the analysis of degraded DNA in disaster victim identification (DVI) and criminal investigations, single nucleotide polymorphisms (SNPs) have been recognized as promising markers mainly because they can be analyzed in short sized amplicons. Most SNPs are bi-allelic and are thereby ineffective to detect mixtures, which may lead to incorrect genotyping. We developed an algorithm to find non-binary (i.e. tri-allelic or tetra-allelic) SNPs in the NCBI dbSNP database. We selected 31 potential tri-allelic SNPs with a minor allele frequency of at least 10%. The tri-allelic nature was confirmed for 15 SNPs residing on 14 different chromosomes. Multiplex SNaPshot assays were developed, and the allele frequencies of 16 SNPs were determined among 153 Dutch and 111 Netherlands Antilles reference samples. Using these multiplex SNP assays, the presence of a mixture of two DNA samples in a ratio up to 1:8 could be recognized reliably. Furthermore, we compared the genotyping efficiency of the tri-allelic SNP markers and short tandem repeat (STR) markers by analyzing artificially degraded DNA and DNA from 30 approximately 500-year-old bone and molar samples. In both types of degraded DNA samples, the larger sized STR amplicons failed to amplify whereas the tri-allelic SNP markers still provided valuable information. In conclusion, tri-allelic SNP markers are suited for the analysis of degraded DNA and enable the detection of a second DNA source in a sample.

Assessment of the stochastic threshold, back- and forward stutter filters and low template techniques for NGM

The AmpFlSTR(®) NGM™ kit shows an increased sensitivity compared to previous AmpFlSTR(®) kits, and the addition of a 29th PCR cycle was found to be the major cause for this. During in-house validation, we evaluated whether the increased sensitivity requires elevation of the stochastic threshold (below which alleles are prone to drop out due to low template amplification effects). To determine the stochastic threshold, over 500 false homozygotes were examined and the threshold was set at the rfu value where 99% of the alleles had a peak height below this value. Using 2085 Dutch reference samples, locus-specific stutter ratios were empirically determined and compared with the ones provided by Applied Biosystems. Application of sharp stutter filters is especially important for the analysis of unequal mixtures. To prevent allele calling of 99% of the -1 repeat unit stutters, thirteen stutter ratio filters could be lowered by up to 1.79% and for two loci the stutter ratio filters had to be elevated slightly with a maximum of 0.06%. At all loci +1 repeat stutters were visible for the higher DNA inputs and for lower inputs at the tri-nucleotide repeat locus D22S1045 as well. The overall +1 stutter ratio filter was set to 2.50% and for D22S1045 it was determined to be 7.27%. To find the optimal strategy to sensitise genotyping for low template DNA samples, a comparison was made between enhancing the capillary electrophoresis settings (9kV for 10s) and increasing the number of PCR cycles (29+5 cycles).

Combining results of forensic STR kits: HDplex validation including allelic association and linkage testing with NGM and Identifiler loci

The autosomal short tandem repeat (STR) kits that are currently used in forensic science have a high discrimination power. However, this discrimination power is sometimes not sufficient for complex kinship analyses or decreases when alleles are missing due to degradation of the DNA. The Investigator HDplex kit contains nine STRs that are additional to the commonly used forensic markers, and we validated this kit to assist human identification. With the increasing number of markers it becomes inevitable that forensic and kinship analyses include two or more STRs present on the same chromosome. To examine whether such markers can be regarded as independent, we evaluated the 30 STRs present in NGM, Identifiler and HDplex. Among these 30 markers, 17 syntenic STR pairs can be formed. Allelic association between these pairs was examined using 335 Dutch reference samples and no linkage disequilibrium was detected, which makes it possible to use the product rule for profile probability calculations in unrelated individuals. Linkage between syntenic STRs was studied by determining the recombination fraction between them in five three-generation CEPH families. The recombination fractions were compared to the physical and genetic distances between the markers. For most types of pedigrees, the kinship analyses can be performed using the product rule, and for those cases that require an alternative calculation method (Gill et al., Forensic Sci Int Genet 6:477–486, 2011), the recombination fractions as determined in this study can be used. Finally, we calculated the (combined) match probabilities, for the supplementary genotyping results of HDplex, NGM and Identifiler.

Autosomal SNP typing of forensic samples with the GenPlex TM HID System: Results of a collaborative study

The GenPlex™ HID System (Applied Biosystems - AB) offers typing of 48 of the 52 SNPforID SNPs and amelogenin. Previous studies have shown a high reproducibility of the GenPlex™ HID System using 250-500pg DNA of good quality. An international exercise was performed by 14 laboratories (9 in Europe and 5 in the US) in order to test the robustness and reliability of the GenPlex™ HID System on forensic samples. Three samples with partly degraded DNA and 10 samples with low amounts of DNA were analyzed in duplicates using various amounts of DNA. In order to compare the performance of the GenPlex™ HID System with the most commonly used STR kits, 500pg of partly degraded DNA from three samples was typed by the laboratories using one or more STR kits. The median SNP typing success rate was 92.3% with 500pg of partly degraded DNA. Three of the fourteen laboratories counted for more than two thirds of the locus dropouts. The median percentage of discrepant results was 0.2% with 500pg degraded DNA. An increasing percentage of locus dropouts and discrepant results were observed when lower amounts of DNA were used. Different success rates were observed for the various SNPs. The rs763869 SNP was the least successful. With the exception of the MiniFiler™ kit (AB), GenPlex™ HID performed better than five other tested STR kits. When partly degraded DNA was analyzed, GenPlex™ HID showed a very low mean mach probability, while all STR kits except MiniFiler™ had very limited discriminatory power.

Analysis of 36 Y-STR marker units including a concordance study among 2085 Dutch males

The genotypes of 36 Y-chromosomal short tandem repeat (Y-STR) marker units were analysed in a Dutch population sample of 2085 males. Profiling results were compared for several partially overlapping kits, i.e. PowerPlex Y, Yfiler, PowerPlex Y23, and two in-house designed multiplexes with rapidly mutating Y-STRs. Nineteen Y-STR marker units, of which two are rapidly mutating, reside in at least two of these multiplexes, and for these markers concordance testing was performed. Two samples showed discordant genotyping results and the probable causative base change was revealed by Sanger sequencing. In addition, we encountered concordant, but aberrant genotyping results including one allele with low peak height and several null alleles. For 12 samples, this involved a null allele in two adjacent loci suggesting a large and recurrent deletion as the samples represent three distinct haplogroups. For each marker unit, the allele counts and frequencies are presented, as are the haplotype counts and haplotype diversities for several combinations of markers.

Understanding the visibility of blood on dark surfaces: A practical evaluation of visible light, NIR, and SWIR imaging

Bloodstains on dark surfaces are often difficult to detect due to a lack of contrast. Infrared photography is in many cases a solution as it enhances the contrast between blood and background. Still, on some surfaces bloodstains cannot be visualized. In this study, we investigate why bloodstains on certain surfaces are not detected and how visibility can be improved. Bloodstains on 166 dark fabrics were photographed by four different cameras and for each, the visibility of the bloodstains was scored. The spectral properties of the dark fabrics were examined as well as the properties of the dyes used to colour the textiles. In addition, spreading of the blood within the textile and the roughness of the material were taken into account. In the investigated set of textiles, visibility of blood is mainly determined by the spectral properties of the textile dye. In addition, a high surface roughness of the textile reduces the visibility.

Improved analysis of long STR amplicons from degraded single source and mixed DNA

DNA profiles from degraded samples often suffer from information loss at the longer short tandem repeat (STR) loci. Sensitising the reactions, either by performing additional PCR cycles or increasing the capillary electrophoresis injection settings, carries the risk of over-amplifying or overloading the shorter fragments. We explored whether profiling of degraded DNA can be improved by preferential capturing of the longer amplified fragments. To this aim, a post-PCR purification protocol was developed that is based on AMPure XP beads that have size-selective properties. A comparison was made with an unselective post-PCR purification system (DTR gel filtration) and no purification of the PCR products. Besides a set of differently and serially degraded single source samples, unequal mixtures of degraded DNAs were analysed, in order to extract more genotyping information for the minor contributor without overloading the major component at the shorter amplicons. Purification by the AMPure protocol resulted in higher peak heights especially for the longer amplicons, while DTR gel filtration gave higher peaks for all amplicon sizes. Both purification methods presented more detected alleles, with the AMPure protocol performing slightly better, on average. In conclusion, the in-house developed AMPure protocol can be employed to improve STR profile analysis of degraded single source and (unequally) mixed DNA samples.