Developing a multiplex PCR‐based assay kit for bloodstream infection by analyzing genomic big data

Background In recent years, the incidence of bloodstream infections (BSI) has increased, the composition of pathogenic bacteria has changed, and drug resistance among bacteria has gradually increased due to the widespread use of interventional techniques, broad‐spectrum antibacterial drugs, hormones, and immunosuppressive agents. Here, we have developed a multiplex PCR assay kit for the detection of pathogens (14 Gram‐negative bacteria, 15 Gram‐positive bacteria, and 4 fungi) in whole blood from patients with BSI using five‐color fluorescent multiplex PCR followed by capillary electrophoresis. Our assay exhibits a diagnosis of higher quality and an improved detection rate for common pathogens. Methods A local genome DNA database of 33 pathogenic bacteria was constructed. Next, “Exhaustive” primer search of the full coding sequence of the reference genomes of these bacteria was performed. Panels with minimal interactions between primers and amplicons were selected by random sampling and testing by a recursive algorithm. Primers and Mg2+ concentrations and PCR reaction procedures were optimized to maximize the detection efficacy. Results The LOD of the kit was determined as 100 copies/μl. Using clinical samples, results generated by this kit and regular blood culture method were found to be 95.08% consistent. Additionally, six pathogens which were unidentifiable by blood culture were successfully detected by this kit. Conclusion Our study provided a bioinformatics approach to the challenge of primer design in multiplex PCR, and combined with optimized wet lab practice, a multiplex PCR‐based assay kit for BSI with higher sensitivity and accuracy than blood culture was produced. A local genomic DNA database of 33 pathogenic bacteria (include 14 Gram‐negative bacteria, 15 Gram‐positive bacteria and 4 fungi) was constructed. Next, “Exhaustive” primer search of the full coding sequence of the reference genomes of these bacteria were performed. Panels with minimal interactions between primers and amplicons were selected by random sampling and recursive algorithm execution. Primers and Mg2+ concentrations, and PCR reaction procedures were optimized to maximize the detection efficacy.