Enabling Early Detection of Emerging Diseases and Biothreats
New infectious disease pathogens are emerging at an unprecedented rate and global travel allows them to spread faster than ever before. In fact, several new viruses, possibly thousands of new bacteria, and approximately 2,500 new fungi are discovered each year. Pathogens like Methicillin-resistant Staphylococcus aureus (MRSA), Rifampicin-resistant Mycobacterium tuberculosis (MTB), and theSARS-CoV-2 virus have all newly emerged in just the past few years with devastating impacts. That makes the ability to detect these threats – before they become an outbreak, epidemic, or pandemic and threaten public health – more critical than ever.
Unfortunately, many of the pathogens that pose threats to public health are discovered after their initial impact, as traditional surveillance detection methods are designed with specific targets in mind – you have to know what pathogen you’re looking for. For instance, when you take a flu test, it will tell you if you’ve tested positive for only the flu. But what if researchers aren’t sure what the target is, and they’re just trying to determine what viruses, bacteria, or fungi are in the environment or if all targeted diagnostic assays have failed and physicians are trying to diagnose a fever of unknown origin?
This is where agnostic detection can play an important role. Using agnostic detection technology can help identify any pathogen, whether it’s a known virus, a novel bacteria, or something entirely new, before it becomes a real threat to public health.
Agnostic Detection Methods in Disease and Threat Detection
In today’s world, pathogens can emerge suddenly, evolve rapidly, or even be deliberately bio-engineered. While traditional detection techniques are target specific, agnostic detection methods are unbiased approaches that can detect a broad range of pathogens, such as viruses, bacteria, and fungi, without needing to know the specific threat in advance. This can include uncovering novel pathogens such as zoonotic diseases or even tracking the spread of drug-resistant organisms, catching antimicrobial resistance early in its evolution. Agnostic methods enable detection of these threats quickly, even if they’re previously unknown or potentially designed to evade traditional tests. Read more about the importance of agnostic detection methods in disease and threat detection in “The Power of Agnostic Detection.”
Addressing Unknown Unknowns The concept of unknown unknowns refers to pathogens or threats we aren’t aware of yet, like the well-known phrase, “We don’t know what we don’t know.” Agnostic detection methods serve an important role in this area because they don’t rely on prior knowledge of specific organisms in the way that pre-designed PCR primers and probes do.
Instead, these methods cast a wide net identifying any genetic material present in a sample by screening them against large, curated databases. This means unknown pathogens that may not be on the radar of traditional tests can still be detected. Even novel or engineered pathogens can begin to be characterized via similarities to known reference strains, which gives researchers correlations that can be used as a head start. By staying ahead of the curve, we can respond to emerging threats, whether naturally occurring or manmade, with greater speed and precision.
Agnostic Detection Methods in Pandemic Preparedness Agnostic detection can play a transformative role in pandemic preparedness by enabling early identification of novel pathogens before they spread widely. During the COVID-19 pandemic, for example, sequencing was key in identifying the SARS-CoV-2 virus, tracking its mutations, and improving diagnostic tests. For future pandemics, agnostic methods could provide even earlier warnings by detecting new pathogens before they’re widely recognized. This would allow health authorities to implement control measures even more quickly, preventing a localized outbreak from becoming a global crisis.
Three Challenges Associated with Agnostic Detection Like any emerging technology, there can be challenges in the use of agnostic detection methods.
One of the main challenges associated with use of this technology is the sheer volume of data generated by agnostic methods like metagenomic sequencing. Analyzing this amount of data can require significant computational resources and expertise.
Additionally, there is a risk of detecting harmless or background microbes that can convolute the interpretation of results leading to false positives.
Another limitation is the cost. Though prices for sequencing have dropped significantly, these methods can still be significantly more expensive than targeted approaches.
Despite these challenges, advances in technology and bioinformatics are steadily improving the efficiency, speed, and affordability of agnostic methods. Read more about the challenges and opportunities associated with agnostic detection methods in “Agnostic Detection Challenges and Opportunities.”
The Trade-offs of Sensitivity and Specificity There can also be trade-offs associated with agnostic detection methods. When designing any detection assay, there are two major characteristics to keep in mind: sensitivity and specificity. Sensitivity is the lowest amount of a pathogen that the assay will be able to detect. So, for example, for a test designed to detect HIV, high sensitivity is critical to ensure that no one with the virus goes undetected, especially in the early stages of infection when viral load in a patient might be low. Meanwhile, specificity in the context of assay design refers to what an assay is supposed to detect. For example, a highly specific assay might detect a singular pathogen strain versus a low specificity assay that can detect everything within a given genus. But there are trade-offs between optimizing for sensitivity versus specificity.
Having high sensitivity and high specificity are often mutually exclusive of each other and must be balanced for the best outcome. This means that the assay designer must consider the end use case and what question they’re trying to answer. Is the end use case a clinician who just needs to identify a pathogen to inform a treatment plan? Or is it an epidemiologist who needs to track specific strains? Or even, will this assay be used on its own or as part of a larger panel of tests? The designer must balance these questions to help dictate what level of sensitivity and specificity they will need to incorporate into their design platform.
While the sensitivity and specificity concepts rule traditional molecular and immune assay design, agnostic detection is actually free from those constraints. This is because in traditional detection assays, the designer typically targets singular proteins or singular gene targets, even in multiplexed assay panels. The conservation and/or copy number of the gene target often dictates an assay’s predilection towards specificity and sensitivity, respectively. On the other hand, agnostic detection methods rely on a total sample profile analysis – and are therefore not reliant on specific a priori gene targets.
Here, researchers are able to map an entire pathogen’s profile, whether that’s a genomic sequence, amino acid profile, or otherwise, against a large library of potential targets. Instead of having to determine a single or short list of gene targets or antigens, a question an agnostic platform designer might ask themselves is, “Is this profile type best suited for the detection, identification, or characterization scenario at hand?” For example, amino acid profiles can be really useful in environmental samples with a potentially wide range of bacteria. On the other hand, sequencing is ideal for detecting novel pathogens, tracking mutations, and being able to screen against huge genomic databases.
The Future of Agnostic Detection Some of the exciting developments that we’ve seen in agnostic detection include improvements in portable sequencing technologies such as nanopore sequencing, which can bring agnostic detection into the field for real-time surveillance. Advances in bioinformatics and AI-driven analysis tools are also making it much easier to interpret complex data sets, enabling quicker identification of pathogens.
Additionally, researchers are working on more affordable and accessible platforms, reducing the barriers to agnostic detection methods and resource limited settings. For example, if a researcher is out in the field and they only have access to one platform, having a singular device that can detect many targets is valuable. These innovations are making it possible to deploy agnostic approaches more widely and effectively in both routine public health monitoring and in response to potential biothreats.
However, it isimportant to recognize that there is no such thing as an omniscient test, as 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 outrightly identify unknowns.Something must be in the library for the result output to confidently match. However, as mentioned previously, most unknowns carry key characteristics that hold similarities to known records and can therefore be used as clues to begin full characterization and/or response.Even a truly unknown samplethat is unrelated to any known records could ostensibly carry clues to its pathogenicity or origin.
We typically see smaller libraries or databases that are highly curated with well annotated and labeled items. By comparison, 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. In the world of agnostic detection, the strength and breadth of possible target determinations is fully reliant on the quality and scale of the selected library.
Making Agnostic Detection More Accessible and Scalable Because new infectious disease threats often emerge in low- and middle-income countries of the world where resources are limited, it is important that agnostic detection becomes more accessible and scalable. There are several ways that this can be supported:
Focus on reducing costs and increasing the availability of portable technologies;
Train people in the global workforce to read and understand bioinformatics so that more countries can analyze the complex data sets that the methods generate;
Make investments in infrastructure that include building more global surveillance networks to allow for the rapid sharing of pathogen data across borders; and,
Establish partnerships between governments, the private sector, and international organizations that are all key to making these technologies scalable and adaptable for use in diverse settings.
Because many platforms currently rely on traditional assay detection methods, this means that every time there is a new pathogen that they want to be able to detect, researchers must go through the entire research and development process in order to deploy that particular assay. With agnostic detection methods, once an assay is developed and deployed, researchers don’t have to continuously revisit the process. Each time a new infectious disease or biothreat emerges, that threat can be detected. And while this approach is not an all-in-one panacea, it is a really exciting development and offers critical advantages in health surveillance.
GETTING STARTED AT MRIGLOBAL Contact MRIGlobal for further information about our work to deliver high-quality products and services to enhance the detection, diagnosis, and prevention of infectious disease in the United States and around the world.
To discuss how we can help your project be successful, contact us today.
SIGN UP FOR OUR NEWSLETTER Sign up for the MRIGlobal newsletter! It’s the best way to get the latest updates in the world of applied scientific engineering research delivered directly to your inbox.
