Development and evaluation of Microbe Finder (MiFi)®: a novel in silico diagnostic platform for pathogen detection from metagenomic data

Abstract

Introduction. With expanding demand for diagnostics, newer methodologies are needed for faster, user-friendly and multiplexed pathogen detection. Metagenome-based diagnostics offer potential solutions to address these needs as sequencing technologies have become affordable. However, the diagnostic utility of sequencing technologies is currently limited since analysis of the large amounts of data generated, are either computationally expensive or carry lower sensitivity and specificity for pathogen detection. Hypothesis/Gap Statement. There is a need for novel, user friendly, and computationally inexpensive platforms for metagenome sequence analysis for diagnostic applications. Methods. In this study, we report the use of MiFi® (Microbe Finder), a computationally inexpensive algorithm with a user-friendly online interface, for accurate, rapid and multiplexed pathogen detection from metagenome sequence data. Detection is accomplished based on identification of signature genomic sequence segments of the target pathogen in metagenome sequence data. In this study we used bovine respiratory disease (BRD) complex as a model. Results and Conclusions. Using MiFi®, multiple target bacteria and a DNA virus were successfully detected in a multiplex format from metagenome sequences acquired from bovine lung tissue. Overall, 51 clinical samples were assessed and MiFi® showed 100 % analytical specificity and varying levels of analytical sensitivity (62.5 %–100 %) when compared with other traditional pathogen detection techniques, such as PCR. Consistent detection of bacteria was possible from lung samples artificially spiked with 109–104 c.f.u. of Mannheimia haemolytica .