Back

Metagenomic Next-generation Sequencing of Cerebrospinal Fluid for the Diagnosis of Central Nervous System Infections: A Multicentre Prospective Study

Fan, S.; Wang, X.; Hu, Y.; Shi, J.; Zou, Y.; Zhao, W.; Qiao, X.; Wang, C.; Chin, J. H.; Liu, L.; Qin, L.; Wang, S.; Li, H.; Yue, W.; Zhang, W.; Li, X.; Ge, Y.; Wu, H.; Chen, W.; Li, Y.; Guan, T.; Li, S.; Wu, Y.; Zhou, G.; Liu, Z.; Piao, Y.; Zhang, J.; Ren, C.; Cui, L.; Liu, C.; Ren, H.; Zhao, Y.; Feng, S.; Jiang, H.; Wang, J.; Bu, H.; Guo, S.; Peng, B.; Cui, L.; Li, W.; Guan, H.

2019-06-10 genomics
10.1101/658047 bioRxiv
Show abstract

BackgroundInfectious encephalitis and meningitis are often treated empirically without identification of the causative pathogen. Metagenomic next-generation sequencing (mNGS) is a high throughput technology that enables the detection of pathogens independent of prior clinical or laboratory information.\n\nMethodsThe present study was a multicentre prospective evaluation of mNGS of cerebrospinal fluid (CSF) for the diagnosis of suspected central nervous system infections.\n\nResultsA total of 276 patients were enrolled in this study between Jan 1, 2017 and Jan 1, 2018. Identification of an etiologic pathogen in CSF by mNGS was achieved in 101 patients (36.6%). mNGS detected 11 bacterial species, 7 viral species, 2 fungal species, and 2 parasitic species. The five leading positive detections were varicella-zoster virus (17), Mycobacterium tuberculosis (14), herpes simplex virus 1 (12), Epstein-Barr virus (12), and Cryptococcus neoformans (7). False positives occurred in 12 (4.3%) patients with bacterial infections known to be widespread in hospital environments. False negatives occurred in 16 (5.8%) patients and included bacterial, viral and fungal aetiologies.\n\nConclusionsmNGS of CSF is a powerful diagnostic method to identify the pathogen for many central nervous system infections.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.