Agnostic Detection Challenges and Opportunities

And while the opportunities for its use are immense, it is not without practical challenges – we note two in particular: 

1. The first of these challenges is with library comparison limitations. Agnostic detection fundamentally relies on an established library of knowns to which it can compare the input. While these libraries can be huge with thousands of references, they, by definition, cannot identify complete unknowns such as a novel pathogen. Something has to be in the library for it to match to. However, similarities between the sample and known records such as genus level markers, antibiotic resistance, or pathogenicity markers can therefore be used as clues to begin full characterization and/or response. We typically see smaller libraries or databases that are highly curated with well annotated and labeled items. Larger databases often have lower quality input labeling or even mislabeling. Even the large library housed at the National Center for Biotechnology Information (NCBI), which is a cornerstone of the biological research community, is known to have flawed inputs.

2. 

   Another challenge is in interpretation of agnostic method detection data. The sheer volume of data generated by methods like metagenomic sequencing means that analyzing it can be a challenge. This requires significant computational resources and expertise, which is not always readily available. To make this data more accessible, investments in infrastructure and training must be made. This will prepare the global workforce with the infrastructure and education to read and understand bioinformatics and the complex data sets that the methods generate, while encouraging rapid sharing of pathogen data across borders.

Read more about how agnostic detection methods enable emerging disease and biothreat detection in “Agnostic Detection Methods for Pathogens.”