Pathogen quantitative efficacy of different spike-in internal controls and clinical application in central nervous system infection with metagenomic sequencing

Abstract

ABSTRACT Metagenomic next-generation sequencing (mNGS) has been a lack of method for pathogen quantitation. We explored the suitable concentration of T 1 phage [internal control (IC) phage ], Thermus thermophilus (IC T.T ), and artificial DNA sequence (IC DNA ) as mNGS ICs for pathogen quantitation and compared the quantitation efficiency among them. We prepared the simulated cerebrospinal fluid (CSF) samples composed of a pathogen cocktail, containing Staphylococcus aureus , Escherichia coli , Komagataella pastoris , and human cells to test the accuracy, linearity, and interference of IC quantitation. We also collected 15 clinical CSF conducting both mNGS and droplet digital PCR (ddPCR) to further verify the quantification efficacy of IC. In accuracy, the mNGS quantification of pathogen was more precise when the IC was 10 3 and 10 4 CFU/mL Thermus thermophilus or 10 3 and 10 4 PFU/mL T 1 phage with the CV% of pathogens quantification most below 15%. The DNA sequence’ quantification was less accurate with all the CV% of pathogen quantitation above 15%. In linearity, compared to DNA sequence (all R 2 <0.9), 10 3 CFU/mL Thermus thermophilus and 10 4 PFU/mL T 1 phage both have stronger linearity in mNGS pathogen quantification according to linear regression (most R 2 >0.9). In interference, the mNGS quantification was affected obviously by human cell concentrations when the IC was DNA sequence ( P < 0.0001), and the quantification was not interfered when the IC was the 10 3 CFU/mL Thermus thermophilus or 10 4 PFU/mL T 1 phage . Furthermore, we revealed the mNGS quantitation was highly consistent with ddPCR in clinical CSF according to the linear regression ( R 2 = 0.9646, P < 0.0001, k = 0.9362) and Bland-Altman (the bias of average difference is 4.033, with the 95% confidence interval from − 24.61 to 32.68). Thermus thermophilus and T 1 phages are comparable as mNGS IC in pathogen quantitation and are both superior than artificial DNA sequences. In total, 1,000-CFU/mL Thermus thermophilus as mNGS IC may allow us to reflect directly the variation of pathogens in central nervous system infection patients. IMPORTANCE Metagenomic next-generation sequencing (mNGS) has been used broadly for pathogens detection of infectious diseases. However, there is a lack of method for the absolute quantitation of pathogens by mNGS. We compared the quantitative efficiency of three mNGS internal controls (ICs) Thermus thermophilus , T1 phages, and artificial DNA sequence and developed the most applicable strategies for pathogen quantitation via mNGS in central nervous system infection. The IC application strategy we developed will enable mNGS analysis to assess the pathogen load simultaneously with the detection of pathogens, which should provide critical information for quick decision-making of treatment as well as clinical prognosis.