Probe-based metagenomic pathogen detection: advancing laboratory capacity for complex diagnosis.

Probe-based pathogen enrichment, followed by NGS, is a promising tool for complex diagnosis, overcoming traditional challenges of shotgun metagenomics, namely small microbial/human genetic material ratio and demanding computational resources. Here, we assessed the combined detection performance of two Illumina probe-based panels, the Respiratory and the Urinary Pathogen ID panels (RPIP and UPIP), using 99 clinical samples of 15 different matrices (e.g., cerebrospinal fluid, plasma, serum, urine, swabs, biopsies, etc.) available from Portuguese National Reference Laboratories. This sample set involved 114 "PCR-positive hits" (Ct values range of 9.7-41.3; median of 28.4) for 52 non-redundant human pathogens. For a more detailed bioinformatics assessment, as a complement of the Illumina turnkey solution (Explify), we applied an extended version of our INSaFLU-TELEVIR(+) metagenomics pipeline. Whereas Explify analyses resulted in an initial detection frequency of 73.7% (84/114), the subsequent application of INSaFLU-TELEVIR(+), including taxonomic classification followed by confirmatory read mapping, enabled an overall detection proportion of 79.8% (91/114) of the PCR-positive hits. This translated into a detection rate increment from 54.3% (19/35) to 65.7% (23/35) for bacteria, and from 85.3% (58/68) to 89.7% (61/68) for viruses. The implemented workflow was also very satisfactory for samples with qPCR Ct values above 30, with an overall detection frequency of 71.8% (28/39) when compared with the 92.0% (46/50) observed for those with Ct ≤ 30. In summary, this study validated and established a pioneering approach at the Portuguese National Institute of Health to support clinicians in complex diagnosis, contributing to advance diagnostic capabilities toward a more informed clinical decision and potential improvement of infectious disease outcomes.