PANDORA Biosurveillance Terms and References

I  

Appendix - Key Terms and Concepts

a. Open-Source Intelligence (OSINT) – The collection and analysis of publicly available data, such as news reports, social media, and government publications, to identify emerging health threats.

b. Ontologies– Structured systems that categorize and link concepts, enabling AI to recognize disease patterns and anomalies in health data. In PANDORA-X, ontologies will help detect unusual health signals by comparing new data to known patterns.

c. Retrieval-Augmented Generation (RAG) is an AI technique that improves accuracy by retrieving relevant external information before generating a response. It combines a search step with a language model, enabling real-time, context-aware answers grounded in up-to-date or domain-specific data.

d. Natural Language Processing (NLP) is a branch of AI that enables computers to understand, interpret, and generate human language. It allows machines to process text or speech in a way that mirrors human communication, making it possible to extract meaning, sentiment, and intent from language.

e. Ontology Development for Anomaly Detection. PANDORA-X will use AI-powered ontologies to detect emerging disease threats efficiently. An ontology is a structured system that organizes knowledge into categories and relationships. In the context of disease detection, it helps AI make sense of vast amounts of health data by linking symptoms, environmental factors, and known disease outbreaks. For example, if AI detects an increase in reports about fever and respiratory issues in a specific region, it can compare this pattern to historical outbreaks and flag it as a potential emerging threat. By continuously refining these connections, PANDORA-X’s ontology system ensures that AI can detect not just known diseases but entirely new and unexpected health threats, making the surveillance system highly adaptable.

f. Next-generation sequencing (NGS) is a high-throughput method that allows scientists to rapidly read and analyze large amounts of DNA or RNA. It enables the simultaneous sequencing of millions of genetic fragments, making it a powerful tool for detecting pathogens, studying genetic variation, and understanding complex biological systems.

g. Federated Learning – A machine learning approach that allows AI models to be trained across multiple locations without sharing raw data, ensuring data privacy while improving detection accuracy.

h. Digital epidemiology is the use of data from digital sources—like search engines, social media, and mobile devices—to monitor, model, and predict disease trends in real time. It complements traditional public health surveillance by providing faster, often broader insights into how diseases emerge and spread.

i. Pathogen-agnosticapproaches do not rely on knowing which specific pathogen to look for in advance. Instead, they scan broadly for any signs of infection or biological anomalies, enabling the detection of both known and unknown threats.

j. Anomaly Detection – A method where AI identifies unusual patterns in health data that may indicate emerging disease outbreaks or biological threats.

k. Isothermal amplification is a method of amplifying DNA or RNA at a constant temperature, eliminating the need for thermal cycling like in PCR. It's faster and more portable, making it ideal for field-based or point-of-care diagnostics. MIRA amplification(Multienzyme Isothermal Rapid Amplification) is a specific type of isothermal amplification that uses multiple enzymes to rapidly and sensitively amplify genetic material. It's particularly useful for detecting pathogens in real-time and in low-resource settings. 

l. Hybrid capture sequencing is a method used to selectively isolate specific DNA or RNA sequences from a sample before sequencing. It uses short, custom-designed probes that "capture" only the regions of interest—such as genes from pathogens—allowing for deep, focused sequencing while filtering out irrelevant genetic material. This approach increases sensitivity and efficiency, especially when targeting rare or low-abundance sequences.

m. Infinitely Multiplexed Detection – A diagnostic capability that allows a single test to detect hundreds or thousands of pathogens at once, improving efficiency and cost-effectiveness.

n. Predictive Analytics – The use of AI and statistical modeling to analyze data trends and predict future disease outbreaks before they occur.

o. LexisNexisis a powerful data aggregation and analytics platform that provides access to a vast repository of news articles, legal documents, public records, and other structured and unstructured data sources. In the context of biosurveillance, it can be used to monitor global health trends and detect emerging threats by analyzing open-source reports and publications in real time.

p. Biosurveillance Nodes – Field-deployed locations equipped with diagnostic tools and AI-driven monitoring systems to detect and track infectious diseases in real-time.

q. Denied areas refer to regions or countries where access to reliable information is limited due to government censorship, restricted data sharing, lack of transparency, or geopolitical constraints. These areas often pose challenges for traditional surveillance systems, making it difficult to detect emerging health threats or biological risks in a timely and accurate way.

References:

1. Parkinson S, et al. 2024. Data Collection and Sharing for Pathogen Surveillance: Making Sense of a Fragmented Global System. Rand Health Q. 2024 Mar 4;11(2):4. PMCID: PMC10911754.

2. Gemler BT, et. al. 2023. UltraSEQ, a Universal Bioinformatic Platform for Information-Based Clinical Metagenomics and Beyond. Microbiol Spectr. Jun 15;11(3): e0416022.PMCID: PMC10269449.

3. Wong F, et al. 2023. Leveraging artificial intelligence in the fight against infectious diseases. Science. 381(6654):164-170. https://www.science.org/doi/10.1126/science.adh1114 

4. Judson SD, et al. 2024. Modeling zoonotic and vector-borne viruses. Curr Opin Virol. Aug; 67:1 01428. PMC11292992. https://pubmed.ncbi.nlm.nih.gov/39047313/ 

5. MacIntyre CR, et al. 2023. Artificial intelligence in public health: the potential of epidemic early warning systems. J Int Med Res. Mar;51(3): 3000605231159335. Erratum in: J Int Med Res. May; 51(5):3000605231178098. PMCID: PMC10052500.

6. JHU/APL’s BIOME Integrates data from multiple sources to create a picture of emerging biothreats & provides early warning & decision-making. Using AI-enhanced surveillance and research data to create an information-rich environment. https://www.jhuapl.edu/work/mission-areas/global-health/facilities

7. Reed GM, et al. 2025. Evaluating Nanopore Sequencing as a Respiratory Virus Diagnostic Tool for the Prehospital Setting. Mil Med. Feb 15: usaf046. PMID: 39953827. 

8. Muggy SD, et al. 2024. Modern Biosurveillance Methods: A Lay Introduction to Effective Use of Nucleic Acid Methods. Rand Health Q. Dec 10;12(1):13; PMCID: PMC11630095.

9. Rancati S, et al. 2024. Forecasting dominance of SARS-CoV-2 lineages by anomaly detection using deep AutoEncoders. Brief Bioinform. Sep 23;25(6): bbae535. doi: 10.1093/bib/bbae535. PMID: 39446192; PMCID: PMC11500442.

10. Gauthier NPG, et al. 2024. Validation of an Automated, End-to-End then Sequencing Assay for Agnostic Detection of Respiratory Viruses. J Infect Dis. Dec 16;230(6):e1245-e1253. PMID: 38696336. https://pubmed.ncbi.nlm.nih.gov/38696336/ 

11. Price, C., et al. 2023. AMAnD: an automated metagenome anomaly detection methodology utilizing DeepSVDD neural networks Front. Public Health, 11: https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2023.1181911/full 

12. Sun ML, et al. 2024. Simple and field-adapted species identification of biological specimens combining multiplex multienzyme isothermal rapid amplification, lateral flow dipsticks, and universal primers for initial rapid screening without standard PCR laboratory. Int J Legal Med. Mar;138(2):561-570. PMID: 37801116.

13. Obiso, R. 2002. A novel approach to detecting gene expression. Genomics and Proteomics. 2(3):46-47.

14. Obiso R, et al. 2004. Digene Corporation. Pharmacogenomics. Jan;5(1):129-32. doi: 10.1517/phgs.5.1.129.25678. PMID: 14683424.

15. Hiraoka S, et al. 2025. Probe Capture Enrichment Sequencing of amoA Genes Improves the Detection of Diverse Ammonia-Oxidising Archaeal and Bacterial Populations. Mol Ecol Resour. Apr;25(3):e14042. PMCID: PMC11887609.

16. Jagadesh, et al. 2023. Mapping Global Bushmeat Activities to Improve Zoonotic Spillover Surveillance by Using Geospatial Modeling. Emerging Infectious Diseases, 29(4), 742-750. https://doi.org/10.3201/eid2904.221022

17. Pfeffer DA, et al. 2018. malariaAtlas: an R interface to global malariometric data hosted by the Malaria Atlas Project. Malar J. Oct 5;17(1):352. PMC6173876. https://data.malariaatlas.org/maps?layers=Malaria:202206_Global_Pf_Parasite_Rate

18. Tajabadi , M. et al. 2024. Privacy-preserving decentralized learning methods for biomedical applications. Computational and Structural Biotechnology Journal. Vol 23: 3281-3287. https://doi.org/10.1016/j.csbj.2024.08.024 

19. Shokr A, et al. 2021, Mobile Health (mHealth) Viral Diagnostics Enabled with Adaptive Adversarial Learning. ACS Nano. Jan 26;15(1):665-673. PMCID: PMC8299938.