Brandon Saint-John

Rapid, reagent-free pathogen-agnostic diagnostics that can be performed at the point of need are vital for preparedness against future outbreaks. Yet, many current strategies are pathogen-specific and require several reagents. We present hyperspectral sensing, using light to non-invasively measure the composition of several molecules to form a spectral signature, to overcome these barriers. To generate these spectral signatures, we present the ProSpectral TM V1, a novel, miniaturized hyperspectral platform with high spectral resolution with two mini-spectrometers. Furthermore, we developed state-of-the-art ML pipelines for near real-time analysis of spectral signatures in saliva samples. We found that we could accurately identify SARS-CoV-2 infection status in double-blinded saliva samples and demonstrate 100% accuracy on a hold out test dataset. To our knowledge, this establishes the fastest hyperspectral diagnostic platform and in a small form factor, and executable with liquid samples, without ligands or reagents, all while maintaining PCR level specificity and sensitivity.

Purpose: variants have been reported in individuals with autosomal recessive hypomyelinating leukodystrophy 18 (HLD18; MIM# 618404). We sought to resolve a 5' +4/+5 splice site variant of uncertain significance found in three individuals with HLD features. Methods: We used next-generation DNA and transcriptome sequencing, cell-based splicing assays, and tandem mass spectrometry to detect and characterize the splice site variant. We then performed RNA structure probing and conventional antisense oligonucleotide screening to investigate molecular mechanisms for potential therapeutic intervention. Results: 5' splice site variant, c.825+4_825+5delAGinsTT (NM_003676.4) was identified in all three participants. Although the gene has been associated with autosomal recessive hypomyelinating leukodystrophy, the variant has not been previously reported in any available databases or literature. We show that the splice site variant: 1) was sufficient to induce exon two skipping in most detected transcripts; 2) resulted in structural changes to the 5' and 3' splice site regions using RNA structure probing; and 3) corresponds to plasma sphingolipid profiles consistent with loss of sphingolipid delta(4)-desaturase activity. Discussion: variant c.825+4_825+5delAGinsTT is pathogenic and suggested a mechanistic model to explain how exon two skipping is induced.

Abstract Background Rapid, reagent-free pathogen-agnostic diagnostics that can be performed at the point of need are vital for preparedness against future outbreaks. Yet, many current strategies (polymerase chain reaction, lateral flow immunoassays) are pathogen-specific and require reagents; whereas others such as sequencing-based methods; while agnostic, are not (as yet) conducive for use at the point of need. Herein, we present hyperspectral sensing as an opportunity to overcome these barriers, realizing truly agnostic reagent-free diagnostics. This approach can identify both pathogen and host signatures, without complex logistical considerations, in complex clinical samples. The spectral signature of biomolecules across multiple wavelength regimes provides rich biochemical information, which, coupled with machine learning, can facilitate expedited diagnosis of disease states, the feasibility of which is demonstrated here. Innovation First, we present ProSpectral™ V1, a novel, miniaturized (∼8 lbs) hyperspectral platform with ultra-high (2-5 nm full-width, half-max, i.e., FWHM) spectral resolution that incorporates two mini-spectrometers (visual and near-infrared). This engineering innovation has enabled reagent-free biosensing for the first time. To enable expedient outcomes, we developed state-of-the-art machine learning algorithms for near real-time analysis of multi-wavelength spectral signatures in complex samples. Taken together, these innovations enable near-field ready, reagent-free, expedient agnostic diagnostics in complex clinical samples. Herein, we demonstrate the feasibility of this synergy of ProSpectral™ V1 with machine learning to accurately identify SARS-CoV-2 infection status in double-blinded saliva samples in real-time (3 seconds/measurement). The infection status of the samples was validated with the CDC-approved polymerase-chain reaction (PCR). We report accuracies comparable to first-in-class PCR tests. Further, we provide preliminary support that this signal is specific to SARS-CoV-2, and not associated with other respiratory conditions. Interpretation Preparedness against unanticipated pathogens and democratization of diagnostics requires moving away from technologies that demand specific reagents; and relying on intrinsic biochemical properties that can, theoretically, inform on all pathologies. Integration of hyperspectral sensors and in-line machine learning analytics, as reported here, shows the feasibility of such diagnostics. If realized to full potential, the ProSpectral™ V1 platform can enable agnostic diagnostics, thereby improving situational awareness and decision-making at the point of need; especially in resource-limited settings – enabling the distribution of newly developed tests for emerging pathogens with only a simple software update. Funding The U.S. Department of Energy, the Defense Threat Reduction Agency, Lawrence Berkeley National Laboratory, Los Alamos National Laboratory, and Pattern Computer Inc. Research in context Evidence before this study Our inability to quickly and effectively deploy and use reliable diagnostics at the point of need is a major limitation in our arsenal against infectious diseases. We searched PubMed and Google Scholar for articles published before May 2024 in English applying hyperspectral sensing technologies of pathogen detection with terms, “hyperspectral,” “pathogens”, and “COVID-19”. Various factors such as speed, sensitivity, availability of reagents, deployability, requirements (expertise, resources), and others determine our choice of diagnostic. Today, diagnosis of infection remains largely pathogen-specific, requiring ligands specific to the target of interest. Indeed, Polymerase Chain Reaction (PCR)-based methods, the gold-standard technology to diagnose COVID-19, are pathogen-specific and have to be re-evaluated with the emergence of new variants. Lateral flow immunoassays, while readily deployable, are associated with lower sensitivity and specificity, and require the development of ligands, which can be time-consuming when addressing unanticipated or new threats. Select pathogen-agnostic methods such as sequencing are evolving and becoming more feasible, but still require sample processing, reagents, cold-chain, and expert handlers - and hence are not (as yet) available for routine point-of-care use. In contrast, the characterization of biochemical signatures across multiple spectral regimes (hyperspectral) can facilitate reagent-free agnostic diagnostics. Yet, many spectroscopic methods are either limited to narrow wavelength ranges; or are too large for use in the point-of-care setting; and may require complex and time-consuming analytics. Added value of this study This manuscript presents a paradigm-shifting miniaturized hyperspectral sensor with embedded machine learning-enabled analytics that can overcome the above limitations, making reagent-free agnostic diagnostics achievable. To our knowledge, this establishes the fastest hyperspectral diagnostic platform (3 seconds/measurement), with no preprocessing and in a small form factor, and executable with liquid (clinical) samples, without ligands or reagents. Our data demonstrates that the sensitivity of this assay is comparable to gold-standard PCR-based assays; and that the signatures are specific to COVID-19 and not associated with influenza and other respiratory pathogens – establishing the truly agnostic nature of the platform. The sensor consists of two embedded spectrometers, covering spectral bandwidth 400-1700 nm, which covers spectral patterns associated with relevant biological moieties. With appropriate data processing, we demonstrate balanced accuracies between 0·97 and 1·0 under a 10-fold cross-validation (depending on the ML/AI algorithm used for prediction). Implications of all the available evidence With the optimization of algorithms and analytical methods and the development of appropriate spectral databases, the ProSpectral™ hyperspectral diagnostics platform can be a flexible tool for rapid, reagent-free pathogen-agnostic detection/diagnosis of disease at the point of need, which can be a disruptive force in our preparedness to counter emerging diseases and threats.

The continued emergence of pathogens, whether novel, re-emerging, or engineered, poses a persistent global biosecurity and public health challenge. Recent outbreaks, including COVID-19, Lassa fever, Marburg virus, mpox, and avian influenza, underscore the urgent need for robust systems that enable rapid surveillance, early diagnosis, and timely countermeasures before widespread human transmission occurs. In this article, we focus on early detection technologies and systematically evaluate current diagnostic and sensing modalities. We highlight sequencing and spectroscopy as two complementary approaches capable of providing broad, agnostic detection and rich biological insight. Our analysis emphasizes that scientific innovation alone is insufficient: effective preparedness also requires improved data curation, integration, and sharing to build AI-ready resources that accelerate future responses. We argue for coordinated advances in both technological capabilities and supporting infrastructure to enable the rapid identification and characterization of emerging pathogens and to fully leverage modern science against evolving infectious threats.

There is an urgent need for truly reagent free and agnostic sensing at the point of need for biological and chemical targets. Indeed, the COVID-19 pandemic has re-emphasized the significant need for truly agnostic diagnostics at the point of need. Whereas highly targeted and tailored diagnostics can help us address a known and anticipated threat, they do not prepare us against the next emerging outbreak. To address the significant challenge of developing truly agnostic diagnostics that can improve situational awareness and guide decision making, our team has looked to spectroscopy as an inspiration for the sensing modality; and innate immunity as an inspiration for the biological process and signatures. All biological molecules emanate spectral signatures along the electromagnetic spectrum, integral information from which can provide insights into processes or signatures more effectively. Therefore, we propose the use of hyperspectral sensing as a modality for accurate diagnosis and detection of signatures. With regards to the biology - innate immunity is a broad, agnostic pathogen sensing strategy that allows for the early recognition of all pathogens – known and unknown – with an associated response. Our team has developed strategies to mimic this response in the laboratory, culminating with the (ongoing) development of a data-science and machine learning enabled effort to unravel the complexity of the immune recognition (host cytokine and chemokine response), allowing for their ability to inform on categories of pathogens/disease. It is important to integrate this knowledge with a point of contact diagnostic approach in order to be able to translate the data into usable diagnostic information. To this end, we are working on developing a hyperspectral approach that measures these signatures without reagents from saliva samples in a few seconds. Charged by a back-end machine learning/artificial intelligence algorithm, our approach uses Pattern’s ProSpectral Sensor that measures signatures across various realms of the electromagnetic spectrum. Data on developing this sensing modality as a diagnostic in real time – challenges and advantages – will be presented.