Fig. 2. Pathogen detection by mNGS.

A Detection of TB for entire cohort. Blue circle represents definite TBM (i.e., detected by GeneXpert Ultra and/or culture). Pink circle represents probable TBM cases based on clinical consensus but GeneXpert Ultra and/or culture negative. Green circle is TBM detected by mNGS. Five additional cases of TBM were detected in possible TB, indeterminate and Not-TBM groups. B Breakdown of pathogens found in possible TBM and indeterminate groups. C Top two graphs are log10(rPM) for normalized DNA-seq and RNA-seq reads to Streptococcus pneumoniae and Neisseria meningitidis detected in every sample sequenced. Red dots indicate samples with greater than 1 log fold abundance greater than the mean cohort abundance in both the DNA-seq and RNA-seq datasets relative to the remaining cohort. Red dots circled in black indicate samples that were diagnosed as bacterial meningitis during hospital admission in Uganda. Bottom two graphs demonstrate a similar method performed for Epstein-Barr virus and cytomegalovirus as both these viruses can be seen in low abundance in neuroinflammatory conditions. Red dots indicate samples with Epstein-Barr virus or cytomegalovirus with rPM 2 log fold higher than the mean cohort abundance in DNA-seq. D Phylogenetic tree for Wesselsbron virus detected in 1 patient in the possible TBM group. Phylogenetic tree was built with a multiple sequence alignment using the assembled genome and reference genomes obtained from National Center for Biotechnology Information. The best-fitting evolutionary model was picked by ModelTest-NGv0.1.5, and a phylogenetic tree was built using RAxML-ng v0.6.0. Scale bar indicates length associated with 4.0 nucleotide substitutions. TBM tuberculous meningitis, TB Mycobacterium tuberculosis, mNGS metagenomic next-generation sequencing, RNA-seq ribonucleic acid sequencing, DNA-seq deoxyribonucleic acid sequencing, rPM reads per million. Source data are provided as a Source Data file.