2

Exploring the Current State of Pathogen Genomics and Metagenomics in the U.S. Public Health Enterprise

The first session of the workshop explored public health applications of pathogen genomics at the federal and state levels. The panel was moderated by David Blazes, deputy director of global health at the Gates Foundation. Greg Armstrong, independent consultant at Ridgway Consulting, LLC, outlined pathogen genomics applications in U.S. public health. Duncan MacCannell, director of the Office of Advanced Molecular Detection at the U.S. Centers for Disease Control and Prevention (CDC), discussed federal applications of pathogen genomics to public health and programs to accelerate innovation in this area. Ruth Lynfield, state epidemiologist and medical director at the Minnesota Department of Health (MDH), provided examples of pathogen genomics use in state-level public health.

CURRENT STATUS OF PATHOGEN GENOMICS IN U.S. PUBLIC HEALTH

Armstrong provided an overview of public health applications of pathogen genomics. Describing a revolution in nucleic acid sequencing technologies, he noted that associated costs have decreased from $10 million per raw megabase of DNA sequence in 2001 to less than $0.01 per raw megabase of DNA sequence today (National Human Genome Research Institute, 2023). He pointed out that the decreasing cost and increasing automation of high-throughput sequencing for pathogen genomics transformed understanding of epidemiology and microbial evolution within research settings, allowed provision of actionable information in public health settings, and enabled emerging applications in clinical settings. Armstrong emphasized

that pathogen genomics is a young and rapidly evolving field with implications across the spectrum of infectious diseases of public health importance, including improved pathogen subtyping, outbreak detection and investigation, inferring phenotype, developing vaccines and therapeutics, and providing a more granular picture of epidemiology.

Applications of Pathogen Genomics in Public Health

Pathogen genomics is replacing various technologies previously used in public health efforts to subtype infectious disease pathogens, Armstrong explained. Whole genome sequencing (WGS) provides higher resolution than older technologies, though it is also more expensive and time consuming. However, continued innovation is increasing automation and reducing the cost of genomics technologies. Outbreak detection efforts employ pathogen genomics in identifying outbreaks ranging from bacterial foodborne illness to COVID-19 to tuberculosis. The National Institutes of Health integrates foodborne illness data from multiple federal programs into the National Center for Biotechnology Information (NCBI) Pathogen Detection system, enabling public access via a single interface. The system integrates clinical data from CDC’s PulseNet,¹ environmental and food data from the U.S. Food and Drug Administration’s (FDA) GenomeTrakr,² and data from the U.S. Department of Agriculture (USDA). During the COVID-19 pandemic, scientists used genomics analyses to determine the sources of clusters of cases and inform interventions. Armstrong noted that in addition to detecting outbreaks, strain-level genomics knowledge can be used after a disease outbreak is identified to, for example, ascertain disease cases that are attributable to the same outbreak, link the outbreak to its source (e.g., a specific packaged food), elucidate transmission pathways, particularly in rapidly evolving viruses, and provide evidence for or against the presence of transmission within health care settings.

Pathogen genomics is also helpful in determining phenotype, Armstrong noted. Prior to genomics, various subtyping technologies generated phylogeny information regarding the relationships between pathogen isolates, but these methods did not provide data about the isolates themselves. In contrast, genomic sequencing often yields much information on pathogen isolates. For instance, genomics has a high level of accuracy in determining Mycobacterium tuberculosis resistance to first-line antibiotics and is now routinely used for first-line drug susceptibility testing in some settings. He posited that, while the accuracy of susceptibility determina-

___________________

¹ See https://www.cdc.gov/pulsenet/hcp/about/index.html (accessed March 18, 2025).

² See https://www.fda.gov/food/whole-genome-sequencing-wgs-program/genometrakr-network (accessed March 18, 2025).

tion for second-line anti-mycobacterial agents is more variable than for first-line therapeutics, genomics will likely be used similarly for second-line drug susceptibility testing in the future. Another use case is phenotyping applications in malaria control programs, where specific Plasmodium genes are routinely sequenced to determine resistance to antimalarial drugs and to detect histidine-rich protein 2 deletions. Many rapid diagnostics for malaria detect the histidine-rich protein 2 antigen and widespread test-and-treat strategies using these rapid diagnostics are hypothesized to exert selection pressure for deletion of the concomitant gene, allowing undetected disease spread and threatening effectiveness of global control programs (Gatton et al., 2017). Finally, Armstrong noted that HIV, a high-consequence viral pathogen, is tested for antiviral resistance through genotypic assays that sequence the viral genome.

Armstrong outlined uses for genomics in the development of vaccines, therapeutics, and diagnostics for pathogens from recurring threats of seasonal influenza to managing outbreaks of known and emerging pathogens like SARS-CoV-2. The United States uses a “sequence-first” approach to monitor seasonal influenza, which includes sequencing of collected samples and determining what subsequent tests may need to be performed based on sequencing data. He described how genomics provides a more granular understanding of epidemiology for emerging pathogens that require heightened situational awareness. For example, during the COVID-19 pandemic, pathogen genomics informed understanding of variant emergence and vaccine breakthrough. During the summer of 2021, genomics data revealed that the Delta variant was emerging globally and accounted for increased cases; this understanding spared time and resources that otherwise would have been dedicated to investigating the cause of higher case numbers, said Armstrong.

Additional Considerations in Pathogen Genomics

Armstrong outlined additional considerations in use of genomics, including substantial variation across pathogen types, sequencing techniques, and technologies. For example, eukaryotic pathogen genomes are on average 1,000 times larger than bacteria genomes, which in turn are 1,000 times the size of average viral genomes. Different techniques are needed to examine these pathogens. In addition to WGS and targeted sequencing technologies, metagenomic sequencing³ is emerging as particularly important within public health, he stated. Moreover, considerations

___________________

³ Metagenomic sequencing refers to the analysis of the total genomic material from a bulk sample (e.g., identifying all viral genomes from a clinical or environmental sample). See https://www.genome.gov/genetics-glossary/Metagenomics (accessed February 17, 2025).

related to the cost of genomics extend beyond the expense of conducting sequencing, which Armstrong highlighted is likely to continue to decrease with improved automation. Sample collection, bioinformatics, obtaining metadata—i.e., associated epidemiological and clinical data, which is important for making use of the genomics analyses—and data integration and analysis contribute to the cost of applied genomics. He described the science and technologies for integrating genomic and epidemiologic data as being in their infancy but noted that this growing area of “genomic epidemiology” stands to improve effective data use in public health. Armstrong also called for workforce development to strengthen genomics efforts in public health, noting the difficulty in recruiting bioinformaticians trained on current methods and the need to increase genomics training within microbiology and epidemiology programs.

APPLICATIONS OF PATHOGEN GENOMICS IN PUBLIC HEALTH: NATIONAL LEVEL

MacCannell described U.S. national efforts to address a multitude of pathogens, including foodborne pathogens and SARS-CoV-2, and highlighted initiatives to accelerate innovation in surveillance for pathogens and diseases. He stated that the breadth of pathogen genomics and metagenomics applications across public health is vast and evolving. He described CDC’s Advanced Molecular Detection (AMD) program, which works across all infectious disease centers at CDC and with state and local public health departments to improve pathogen detection and characterization. Established by a presidential initiative in the 2014 fiscal year with a $30 million budget, AMD currently operates with a $40 million annual budget to address its broad and complex mission, he noted. With a focus on implementing emerging laboratory technologies, AMD works toward the objectives of improving pathogen detection and characterization, building integrated informatics systems to analyze and transform data into meaningful public health actions, strengthening the technical workforce and improving access to technology and skills, and promoting open data standards, interoperability, and reproducible methods to increase sustainability, accessibility, and usage of pathogen detection approaches.

The Beginning: Foodborne Disease Pathogen Genomics Surveillance

MacCannell described the expansion of genomics capability over the past decade. In 2013 fewer than 10 state public health laboratories had sequencing capacity. Driven by the PulseNet and GenomeTrakr foodborne disease surveillance systems, next-generation sequencing technologies were adopted rapidly, and by 2018 sequencing capacity was available in every

state public health laboratory. Many city and county public health laboratories now have some level of sequencing capacity with applications that extend beyond foodborne disease surveillance, he noted.

A 2013 pilot program that addressed an ongoing outbreak of Listeria monocytogenes from sliced deli meats provided early proof of concept for pathogen genomics surveillance at a national scale. Each year the United States sees approximately 1,600 clinical cases of L. monocytogenes, which may cause devastating infections, particularly among immunocompromised people, expectant mothers, and other vulnerable populations (CDC, 2024b). Thus, use of pathogen genomics for detection of and response to Listeria outbreaks constitutes a tractable, high-impact public health opportunity, he explained. In September 2013, CDC, FDA, USDA Food Safety Inspection Service, and NCBI entered a collaboration with the goal of sequencing all available clinical, environmental, and food isolates to assess the effectiveness and feasibility of WGS for foodborne disease surveillance. This effort engaged state and local public health departments and international public health authorities, leading to rapid and marked improvements in outbreak identification, said MacCannell. Within the first two years of Listeria surveillance using WGS, the number of detected case clusters increased from 14 to 21, signifying that several outbreaks had been identified and addressed that otherwise would have gone undetected or experienced detection delays. Consequently, Listeria-related food recalls and media coverage increased, raising public awareness (Jackson et al., 2016). This capability achieved through application of pathogen genomics enabled more rapid and tailored public health interventions, MacCannell shared. The effects of earlier detection are evident in the increasing number of outbreaks resolved through identification of the contaminated food source. In the first two years of the pilot program, that number rose from one to nine resolved cases per year, and the median number of cases per cluster decreased from six to three during the same period.

Scaling pathogen genomics in disease detection from Listeria to broader applications for different pathogens involves significant logistics and feasibility considerations. MacCannell described variation across federal surveillance networks stemming from differences in objectives for mitigation of different pathogens. The objective of foodborne disease surveillance is generally rapid identification and intervention in multijurisdictional transmission events in the food supply. Objectives for tracking nonfoodborne pathogens, such as tuberculosis or influenza, prioritize analyzing trends over time rather than real-time detection. Much of the bioinformatics and sequencing of foodborne disease pathogens takes place at the local level, whereas other pathogens may require wider distribution of sequencing and bioinformatics activities. Models of data processing also vary for different pathogens. Foodborne disease surveillance uses a centralized model in

which all data pass into a command center for analysis, rapid assessment, and classification. Disease surveillance networks for some illnesses, such as tuberculosis or influenza, use a reference center model, in which sequence data pass through designated central sites that work to identify trends in disease burden, antimicrobial susceptibility, or phenotypic changes that may affect pathogen virulence or protective immunity. Unlike with foodborne disease surveillance, these data are not necessarily used for real-time, large-scale disease or pathogen tracking. MacCannell added that many current models are somewhat decentralized, with different sites performing analyses in a coordinated manner but not necessarily contributing toward a centralized data exchange.

Application to an Emerging Pathogen: SARS-CoV-2 Genomic Surveillance

MacCannell stated that genomic sequencing of SARS-CoV-2 demonstrated a high degree of collaboration between federal, state, and other partners with public health, with U.S. laboratories in different sectors cumulatively generating more than five million sequences from collected samples. In 2020, the first year of the COVID-19 pandemic, sequencing was performed primarily at the state and local levels, with federal laboratories increasing sequencing activities as the pandemic progressed. MacCannell pointed out that SARS-CoV-2 sequences were submitted within a broad range of timelines, and many institutions did not prioritize sequencing until late 2020, when B1.1.7/Alpha was designated a “variant of concern.” He stated that although all state public health laboratories had the capability and in-house staff for sequencing, the ability to quickly pivot to an unknown pathogen was often limited. This illustrates that genomic sequencing and bioinformatics capacity built around routine pathogen surveillance do not automatically translate into readiness for a novel threat. Many of the states that successfully pivoted to SARS-CoV-2 detection and tracking early in the pandemic were recipients of preexisting support from AMD for workforce development, particularly in the areas of bioinformatic regional resources and bioinformatic training leads. He acknowledged that selection criteria for these investments included the presence of well-established sequencing capabilities, so these state programs had strong foundations before receiving this additional support. Nonetheless, investing in a skilled workforce appears to contribute to more highly adaptive genomic capabilities for pathogen detection and monitoring, said MacCannell.

MacCannell also highlighted the value of partnerships in strengthening and accelerating pathogen genomics, noting that in many states, academic and clinical partners submitted SARS-CoV-2 sequences before the public health laboratories did. In 2020, spurred by the COVID-19 pandemic, AMD established the Sequencing for Public Health Emergency Response,

Epidemiology and Surveillance consortium to harness organically occurring sequencing activities for SARS-CoV-2 in the United States and to incorporate these efforts with public health efforts at local, state, and federal levels. The consortium now includes more than 1,800 international scientists from more than 200 organizations, noted MacCannell.⁴

MacCannell explained that monitoring the SARS-CoV-2 pathogen reveals different goals at different levels of the public health system. National public health focused on generating baseline data for detection and monitoring of viral variants, conducting reference characterization and risk assessment, tracking mutations with implications for diagnostics, vaccines, and therapeutics, guiding resource prioritization and strategy (e.g., distribution of therapeutics and diagnostics), and contributing to a global picture of circulating viral variants. Although these issues are also pertinent at other levels of public health, primary goals for state and local efforts earlier in the COVID-19 pandemic focused on developing tactical responses to investigate outbreak clusters, conducting reference characterization, and conducting contact tracing for infection control, MacCannell stated. MacCannell remarked that the use of metagenomics in public health is largely driven by specific use cases, is typically limited in scale, and often remains in the pilot phase or is secondary to other technologies. For example, in 2012—before the AMD program—CDC and local authorities in Uganda identified an atypical outbreak of yellow fever using a pathogen-agnostic metagenomic approach, which informed a mass vaccination campaign. Other applications of metagenomics remain exploratory, such as efforts to determine whether foodborne disease pathogens can be sequenced directly from stool samples, whether useful genotype and antimicrobial resistance (AMR) profile for tuberculosis can be determined from sputum samples, and whether the transmission of other pathogens (bacteria, fungi) or transposable elements results in public health impacts. MacCannell highlighted the main challenges for metagenomics applications: cost, data complexity and interpretability, timeliness, reproducibility, and regulation of technology use.

Accelerating Innovation in Pathogen Genomics

To fill knowledge gaps and promote innovation, AMD has launched 45 projects through Broad Agency Announcements since 2020, MacCannell noted. Initially, Broad Agency Announcements awards focused on SARS-CoV-2 detection and monitoring capabilities, epidemiologic investigations, and transmission diagnostics. This focus has since broadened to include

___________________

⁴ See https://www.cdc.gov/advanced-molecular-detection/php/spheres/index.html (accessed November 28, 2024).

pathogen-agnostic and multipathogen detection approaches. Additionally, AMD is addressing workforce needs via the Association of Public Health Laboratories–CDC Infectious Disease Bioinformatics Fellowship Program. MacCannell shared that 10 cohorts totaling 66 fellows have completed the fellowship between 2014 and 2023, with 71 percent of fellows remaining in public health to date.⁵ Currently, Association of Public Health Laboratories and CDC are reconfiguring this program to improve recruitment and incorporate a stronger focus on genomic epidemiology. In addition, CDC has funded the Dr. James A. Ferguson Emerging Infectious Diseases Research Initiatives for Student Enhancement Fellowship Program to expand workforce training opportunities, said MacCannell.

Launched in late 2022, the CDC Pathogen Genomics Centers of Excellence (PGCoE) network is designed to foster innovation and technical capacity in pathogen genomics, molecular epidemiology, and bioinformatics, said MacCannell. The PGCoE network has established five U.S. centers, each featuring multiple academic partners and led by a state or local public health department or laboratory. The program creates opportunities for networking at local and state levels to drive applied innovation and research and development (R&D) in public health and to ensure that collaborative networks for outbreak and pandemic response are in place. For example, the creation of standardized institutional review board applications and data-use agreements prevent the need to build these tools retroactively during a crisis. He stated that PGCoEs increase resilience and surge capacity and foster collaboration between public health and academia, which includes labs at research-intensive medical centers that analyze local clinical samples, to acquaint the future workforce with public health programs and careers. MacCannell added that CDC is exploring opportunities for synergy and interconnection between PGCoEs and centers of excellence hosted by other federal agencies and academic groups.

MacCannell noted that addressing informatics needs across public health, from ensuring privacy and data compatibility to building infrastructure for information technology and data access, is challenging since every federal agency and each state and local public health entity has its own policies, governance, and preferences in terms of software and vendors of cloud data storage. To address information technology and access issues, a new initiative called the AMD Platform, which features a modular set of cloud-based bioinformatics services with reproducible code, has been made available to state and local public health departments. According to MacCannell, the AMD Platform can increase the stability, accessibility, and availability of bioinformatics tools.

___________________

⁵ As of the workshop on July 22–23, 2024.

APPLICATIONS OF PATHOGEN GENOMICS IN PUBLIC HEALTH: STATE LEVEL

MDH is one of the CDC’s five PGCoEs. Lynfield described MDH’s use of pathogen genomics to address foodborne pathogens, SARS-CoV-2, and group A Streptococcus. Lynfield explained that state-level public health departments and laboratories are responsible for diagnostic testing of specimens, pathogen isolate characterization, infectious disease surveillance, investigation of clusters and outbreaks, and applied research to improve population health. Epidemiologists use data from disease surveillance and other sources to study patterns of disease incidence and develop and evaluate prevention and control strategies. Lynfield underscored the importance of collaboration between epidemiologists, laboratory scientists, and bioinformaticians within state public health efforts.

Foodborne Pathogens

Lynfield characterized the advantages of pathogen genomics, which allows epidemiologists and researchers to determine the relatedness of strains, identify virulence factors, and detect AMR. Before the advent of WGS, she noted, scientists used pulsed-field gel electrophoresis (PFGE) to glean information about pathogen genomes. In a study comparing PFGE and WGS methodologies to examine cases of Salmonella enterica, WGS provided higher resolution, doubled the number of detected case clusters, reduced the number of cases per cluster, decreased the duration of the cluster, and allowed epidemiologists to focus resources on investigating causes of the outbreaks (Rounds et al., 2020). Lynfield outlined a 2024 Minnesota case of L. monocytogenes in a pregnant woman that caused fetal demise at 29 weeks gestation. An interview using the standardized CDC Listeria Initiative questionnaire revealed exposure to a certain cheese purchased from a large national retailer. However, WGS indicated that the woman’s isolate was related not to isolates from a known national outbreak related to this type of cheese but rather to environmental cheese isolates from Ecuador and to a 2017 Oregon case patient who had consumed the same type of cheese in Ecuador. A second interview revealed that the woman had purchased Ecuadorian cheese from an online store. Sequencing of cheese samples from the store revealed a genetic relationship to the Minnesota case. In response to this finding, FDA issued an import alert and conducted an investigation, while MDH conducted public awareness efforts via social media.

SARS-CoV-2

Lynfield described how MDH conducted SARS-CoV-2 sequencing during the COVID-19 pandemic in response to four outbreaks in long-term care (LTC) facilities, two outbreaks in correctional facilities, and two outbreaks in meat processing plants. Sequencing indicated that isolates obtained from within one LTC facility were related to each other, while in another LTC facility “outbreak cases were associated with two distinct sequences” that were co-circulating (Lehnertz et al., 2021). Further investigations found that one of the isolates from the second LTC setting matched those from a correctional facility and were linked to a household where employees from both the second LTC facility and the correctional facility resided. Both individuals were symptomatic at the same time and tested positive for COVID-19 in mid-May 2020. Lynfield highlighted that the outbreak in the LTC facility started at the end of April and the correctional facility outbreak began in mid-May, and this investigation confirmed COVID-19 had spread between these facilities. In contrast to LTC and correctional facilities that contained highly related strain clusters, sequencing performed on isolates circulating in meat-processing plants featured substantial variation, including multiple distinct strains. A phylogenetic tree of these isolates included multiple clusters with viral strains found both in meat-processing plant employees and community members, Lynfield noted. These shared strains indicate multiple introductions of SARS-CoV-2 to the plant—with employees becoming infected in the community and potentially transmitting the virus back into the community—rather than widespread transmission within each plant, Lynfield explained.

Group A Streptococcus

Lynfield highlighted an outbreak of invasive group A Streptococcus that continued for three years in an LTC facility, noting the high morbidity and mortality rates typically associated with this pathogen in LTC residents. In collaboration with CDC, MDH conducted WGS and PFGE analyses on isolates collected from the facility in question and from other LTC facilities. PFGE revealed there were four case clusters across all LTC facilities examined, and WGS indicated that isolates from different facilities were related to one another. Further investigation identified a wound care provider who treated residents across these LTC facilities. Once this individual was evaluated and received additional training on infection prevention, the outbreaks ended, Lynfield stated.

Lynfield described public health as a collaborative sport and underscored the value of partnerships between CDC and academia. As part of the PGCoE, MDH collaborates with academic facilities on projects such as

examining the spread of carbapenem-resistant Enterobacterales in people and companion animals, developing a tool to predict Salmonella outbreaks, and determining sources of foodborne illness outbreaks.

DISCUSSION

Addressing Variability in Capacity for Genomics Surveillance of Infectious Diseases Across the United States

Blazes asked about challenges in ensuring that all 50 states have comparable genomics capacity for pathogen detection and monitoring within a federal system. MacCannell commented on the variations in capacity across the public health system—as evidenced in multiple jurisdictional outbreaks—and on challenges that are not necessarily related to public health. In many cases, the limiting factor in public health capabilities is not the process of sequencing a pathogen but the decision-making process and related methodology and logistics, he remarked. For example, pathogen-related contextual data are often contained in analog sources (e.g., clinical notes, hospital information systems) that are not structured for efficient access and analysis. This limits the ability to associate epidemiologically relevant metadata with genomic samples within a useful timeframe. Additionally, interjurisdictional data-sharing issues often add complexity to outbreak response when multiple entities are involved.

Lynfield described the Association of Public Health Laboratories–CDC Infectious Disease Bioinformatics Fellowship Program as essential to MDH’s ability to expand beyond sequencing of enteric pathogens to explore other areas. She also underscored the value of additional support to facilitate collaborations between states, through initiatives like the PGCoE and other Broad Agency Announcements awards, and the State Public Health Bioinformatics Group (a consortium that fosters grassroots collaboration). Different states are unlikely to have equivalent resources, she said, but collaboration enables each state to share knowledge regarding its areas of expertise. Lynfield called for greater investment in genomic surveillance of infectious diseases, given that the ability to detect and stop an outbreak or prevent it from occurring will yield long-term financial savings for society. Armstrong remarked on the benefits of shifting from a traditional model—where CDC is at the center—to a network model in which state health departments serve as centers of expertise and interact more effectively with one another. The network model also better aligns with the U.S. system, in which most regulatory authority in health is held at the state level, he added. However, although local control offers advantages, it also contributes to variability across U.S. public health capacity, Armstrong acknowledged. He pointed to the U.S. food system as an example of

a system whose national scope warrants a strong federal role and expressed support for continued national investment.

Challenges in Developing Sequencing-Based Diagnostics

One participant asked how to foster greater application of sequencing to diagnostics, given that a sequencing-based diagnostic technology has been approved by FDA for HIV that also detects drug resistance. In response, MacCannell described genomics as upending the status quo of pathogen-specific public health efforts. Although pathogen genomics applications use domain-specific expertise and sampling, they also require cross-domain capacity for implementation, which is challenging to achieve in a complex public health system, he explained. MacCannell added that implementing regulatory authority over sequencing-based tests is complicated because of issues involving differences in bioinformatic pipelines, test reproducibility, and reliability of a complex test with intended deployment in many different subpopulations. However, he noted that the regulatory community is discussing how to establish gold standards, create high-quality databases, and expand validation capacity that would facilitate future development of sequencing-based diagnostics.

Pathogen Genomics Applied to Outbreak Prevention

A participant asked about the potential to expand genomics surveillance for infectious diseases from outbreak response to outbreak prevention via early pathogen identification. Armstrong replied that much public health sequencing is performed for these objectives. For example, PulseNet continuously conducts analyses to identify foodborne disease pathogens. Similarly, nationwide sentinel sites constantly conduct sequencing of Streptococcus pneumoniae and seasonal influenza viruses. Armstrong remarked that outbreak preparedness has generally been bolstered with readily available pathogen genomic data and expanded interaction between public health and academia, thereby creating a more dynamic outbreak response system. MacCannell described achieving early warning for outbreaks as a main goal for pathogen genomics. National SARS-CoV-2 “nowcasts” are an example of using current and recent laboratory tests to predict trends in disease incidence in different areas, and researchers have explored using large amounts of sequencing data to predict global trends for a given pathogen within a complex immunological landscape, he explained. Harmonized data sources from wastewater and other modalities are likely to improve modeling of features that cause pathogen emergence, said MacCannell.

Lynfield highlighted a collaborative project between MDH and University of Minnesota that examined spread of respiratory illness pathogens within rural communities with limited health care access. A tabletop exer-

cise explored potential public health responses, including follow-up measures and epidemiologist mobilization, in the event that a sample collected from ill community members indicated the presence of highly pathogenic avian influenza A. Lynfield noted the multidisciplinary nature of outbreak prevention, highlighting that CDC’s Center for Forecasting and Outbreak Analytics has established a network of academic partners and state health departments and has also began collaborating with PGCoEs.

Objectives for Genome Sequencing

An attendee commented that WGS rarely involves sequencing of the entire genome but is instead understood to imply deep sequencing with a focus on specific elements of interest to researchers. The participant noted that partial sequencing of the genome may fail to identify novel elements that have developed in known pathogens (e.g., the “unknown unknowns”), which can often be important to understanding emerging health threats such as COVID-19 or mpox. Armstrong replied that for most public health applications, specific objectives drive sequencing efforts and inform decisions regarding the necessary technology. This often does result in focus on and use of specific parts of the pathogen genome. He noted that researchers do conduct additional sequencing as appropriate. MacCannell emphasized the importance of alignment between methodology and objective. For instance, identifying novel elements in a pathogen genome may not be aligned with objectives for national, regional, or state-level routine public health genomic surveillance for infectious diseases. For decades, pathogen classification and disease epidemiology have used single gene identification, and studying the characteristics of individual genes can also identify extra chromosomal sequences and other segments relevant to public health decisions. Scaling is dependent on the objective, he remarked. The sequencing modality and bioinformatic pipeline will change in response to the setting, context, and specific research question of interest.

Data sharing can make data more actionable and meaningful, said MacCannell, adding that a commitment to depositing sequences in the NCBI database for broad availability enables active R&D efforts. Armstrong stated that pathogen genomics has accelerated long-standing collaboration between CDC and academia. Collaborative partnerships expand the capacity to realize WGS benefits.

Federal Agency Collaborations for Genomic Surveillance of Infectious Diseases

Participants asked about collaborations with the U.S. Department of Veterans Affairs, the U.S. Department of Defense, and the U.S. Department of Homeland Security, particularly in terms of data sharing. MacCannell

noted there has been intermittent collaboration with these departments around detection and monitoring for specific pathogens, diseases, and AMR. These departments are discussing how to improve interconnection and integration of these disease and pathogen surveillance systems beyond the current pathogen-specific collaborations, he added. However, MacCannell noted that a barrier has traditionally separated Department of Defense from civilian agencies, which may limit translation of innovative technologies between the two sides. Lessons learned from routine pathogen surveillance at domestic and global levels highlight opportunities for improved communication and pathogen-agnostic methods, he said. These opportunities include developing flexible technologies and approaches that can be applicable to a wide range of uses and addressing workforce challenges, technical issues, and lab capacity limitations. Too often, federal agencies build parallel workstreams in lieu of collaboration, and the possibility of collaboratively developing and using common tools is an open opportunity to accelerate innovation, said MacCannell. A participant noted that the Department of Veterans Affairs Science and Health Initiative to Combat Infectious and Emerging Life-Threatening Diseases—a sizable biorepository of specimens from veterans—is working to expand the demographic representation of specimens and is open to collaborations with external agencies.

State-Level Collaboration and Data Sharing

Participants asked about the formation of PGCoEs and about approaches state governments, public health entities, and other partners might take to overcome data-sharing restrictions. Armstrong replied that to ensure collaboration, the PGCoE network—modeled on CDC’s Emerging Infections Program created in the mid-1990s—requires state health departments and local academic institutions to submit joint proposals. He briefly outlined confidentiality considerations regarding data sharing, noting that notifiable disease data collected in the interest of public health, unlike research data, are often gathered without individual permission and subject to use limitations, as health departments have a duty to prioritize confidentiality. This poses challenges to sharing data outside of health departments. Although genomic data can often be made public without compromising confidentiality, in some cases data sharing could jeopardize confidentiality, he explained. For instance, much of the sequence metadata are not made public for Listeria cases because the relative rarity of those cases makes them potentially identifiable. Armstrong noted the challenges of determining data consolidation needs at the federal level and of establishing agreements in advance to enable rapid, effective data sharing. MacCannell remarked that much data in health care settings have been siloed in accordance with specific applications and by pathogens and thus have not

been captured in standardizable electronic systems. He noted that there may be opportunities in public health data modernization efforts to use artificial intelligence (AI) to classify and incorporate data in real time from poorly structured data sources.

Benefits of Increased Case Cluster Identification

Another attendee asked about the implications of discovering more case clusters using WGS compared with using PFGE, specifically on resources needed for outbreak investigation. Lynfield replied that WGS both identifies more case clusters and enables more efficient use of resources by focusing investigative efforts on disease outbreaks. She said PFGE was more prone to identifying incident cases that are unrelated to an outbreak, creating “noise” in disease outbreak investigations. In contrast, WGS and multilocus sequence typing increase outbreak investigation efficiency in detecting connections between cases. At the national level, this enables multistate outbreaks to be more readily identified, Lynfield added.

This page intentionally left blank.