Drawing a Line in the Sand: Adjudication of Targeted Infectious Disease Next Generation Sequencing Assays

Next generation sequencing (NGS) is a promising technological candidate for future molecular diagnostics due to its inherent capacity for multiplexed and culture-independent testing. Coupling NGS with upfront PCR amplification... [ view full abstract ]

Next generation sequencing (NGS) is a promising technological candidate for future molecular diagnostics due to its inherent capacity for multiplexed and culture-independent testing. Coupling NGS with upfront PCR amplification of unique pathogen nucleic acid signatures greatly enhances the sensitivity, specificity, and signal-to-noise ratio (SNR) for detection of pathogen over host or other environmental genetic material. A robust sample-to-answer scheme for targeted infectious disease assay therefore incorporates sample processing and target amplification, followed by library preparation, sequencing, and analysis. While the benchtop portions of this workflow are well established and under continuous improvement, bioinformatics analysis and final assay adjudication from resulting low-diversity amplicons is currently lacking and plagued by high false positive rates when resolved on Illumina NGS instruments. We therefore developed a bioinformatics pipeline and classification algorithm that accepts sequence reads as input and adjudicates positive, negative, and indeterminate diagnostic outcomes. Reads are first mapped to target sequences and filtered before detailed mapping statistics, including depth of coverage, linear coverage, identity to reference, base quality, and mapping quality, are calculated from the alignments of each target. A two-dimensional threshold applied to the depth of sequencing and the weighted multiplicative product of coverage, identity, base quality, and mapping quality is then used to differentiate diagnostic outcomes. Experimental validation of numerous and diverse mock-clinical Multiplex Target Enrichment NGS (MTE-NGS) West Africa panel datasets run on both the Illumina MiSeq and MiniSeq instruments and analyzed using a MATLAB software prototype shows marked improvement in accuracy and precision of the novel classification algorithm over preexisting methods. This development represents the first bioinformatics method for absolute adjudication of targeted infectious disease NGS assays with future potential for expansion to other target amplification and sequencing technologies and integration into the EDGE Bioinformatics platform for a user-friendly sample-to-answer diagnostic solution.