Group 1
Jennifer Chen
Grade 12
Lynbrook High School, San Jose CA
Massively Parallel DNA Sequencing-Based Human Identification
The forensic DNA backlog represents the accumulation of unanalyzed samples that have the potential to convict the guilty, exonerate the innocent, and protect the general population from repeat offenders, making the reduction of such a backlog critically important.
My project proposed a human identification method based on massively parallel DNA sequencing (MPDS) to process a large number of samples at the same time. However, such an economy of scale depends on large-scale barcoding of the samples that are batched together so that they can later be sorted to determine individual profiles. I therefore designed a linking primer strategy, using a two-stage barcoding process to reduce the total number of PCR primers needed.
I tested the MPDS-based method with the linking strategy on three reference DNA samples (K562, 9947A, and 9948). I also developed a UNIX-based method of determining the DNA profiles since there is no existing software to analyze such a massive number of sequences for forensic purposes.
Of the 11 loci for which complete sequencing reads were obtained, alleles were successfully designated for 10 loci of K562, 7 loci of 9947A, and 9 loci of 9948. Failure to call the second allele for a heterozygous locus was due to an imbalance in read count between the two alleles, but this is an issue of PCR not specific to the linking strategy. The linking strategy was shown to effectively incorporate all oligonucleotide tags, and, as a whole, the cost of the MPDS method compares favorably with that of the current low-throughput method, offering a reduction of up to 75% in consumable expenses and more, due to the potential to cut manual labor.
In conclusion, high throughput and cost reduction of an MPDS-based human identification method enabled by the linking primer strategy hold the potential for addressing the forensic DNA backlog.
