Deciphering the nexus of AMR in One Health using Digital Innovation

A multi-omic AI/ML-enabled data collection tool for AMR detection and surveillance. 

Harry Akligoh is the CEO of the early-stage startup applying for the grant.

Antimicrobial Resistance (AMR) is a significant global health threat, resulting in about 5 million deaths in 2019, with most of these casualties occurring in LMICs. According to a 2023 report, Ghana spends more than US$1 million annually in AMR-associated healthcare costs.

AMR research and surveillance data collection, such as those gathered under the  Fleming Funded initiative in Ghana have primarily been focused on clinical human studies of hospitalized patients. Unfortunately, data from these sources have worryingly been fragmented, limiting their utility. These difficulties continue to be a growing threat, highlighting an urgent need for innovative alternatives to address AMR. 

Addressing this challenge, we conducted a proof-of-concept of our innovative digital solution ( MicroBIS) through a software beta test to detect, manage, and track the community spread of AMR in Ghana. Our approach combines phenotypic and genomic data generated outside the traditional means of AMR data collection ( clinical studies). We utilize samples from wastewater, a key feature of our innovation that presents AMR as a multi-factorial problem between humans, animals, and the environment (One Health). 

Overall, our platform solution will enhance data centralization, ease in data accessibility, which will facilitate more effective AMR research and surveillance.

Our solution caters to various stakeholders in the antimicrobial resistance (AMR) diagnosis, research, and surveillance sectors. These stakeholders include lab technicians, AMR researchers, citizen scientists, ministries of health and governments. In their daily roles, these users engage in identifying (diagnosing), documenting, and making decisions regarding the distribution of AMR for infection prevention and control. Unfortunately, the data generated by these stakeholders are siloed and disconnected resulting in their poor utility and/or accessibility for data driven insights. Our solution creates an interoperable platform that streamlines bacteria identification, and AMR data harmonization from wastewater surveillance and clinical/hospital studies. By enabling access to a comprehensive database, our solution deepens the AMR data pool for asking novel questions and developing actionable strategies for public health interventions and AMR surveillance.

There are many benefits that the public derives from our solution. First, we will publicly disseminate data generated from microBIS. This will be done using an approach we have previously implemented through an interactive public dashboard. 

In addition, using open data-sharing while adhering to data-sharing standards and protocols (eg: HIPAA guidelines), we will make available non-sensitive AMR data to researchers to foster new collaborations and attribution.

Moreover, our platform enhances data centralization which can drive more precise and accurate identification of bacteria in test samples submitted to our platform. 

Our platform also facilitates early detection of microbes that could potentially cause outbreaks, leading to swift outbreak preparedness response strategies. As the COVID-19 has shown, such pandemics can have dire repercussions.

microBIS offers an innovative tool that provides a scalable and cross-data platform for AMR detection and surveillance data management. This facilitates the harmonization of fragmented AMR datasets that currently exists in Ghana, and other LMICs. Such a centralized repository connects various stakeholders in AMR research and surveillance, which will ensure effective collaboration, and equitable access to AMR data, galvanizing innovative research and development.

We therefore expect our solution to have an immediate impact on AMR research leading to much needed inventive technologies to reduce the global burden of AMR. Our expectation is that, over time, scientific research on AMR, policies on AMR surveillance and clinical diagnosis of infectious bacteria will largely depend on data from our platform.

Duplex Bioscience's primary objective is to establish Africa's first comprehensive digital infectious disease database and pathogen DNA sample repository. This ambitious undertaking requires significant investments in physical infrastructure and computational software resources. Our current project on "Decipher Antimicrobial Resistance (AMR)", currently limited to Ghana, has the potential to serve as a crucial foundation for our broader program.

We note the similarity in the challenges of AMR across LMICs, but are also cognizant of the variability in data generated across different geographic locations. Therefore, in the next couple of years, we plan to extend our data collection approach to selected LMICs (such as Nigeria). This will increase the breadth of data used in training our predictive models, and hence improving the accuracy of our predictions, as well as making our platform more generalizable. We will also be able to provide more targeted services to various geographies. 

To gauge the success of our project, we will employ the following key performance metrics:

1. Implementation of a fully functional web-based platform for AMR detection, accompanied by a dashboard for AMR distribution mapping.
2. Establishment of a centralized and interconnected AMR database.
3. Assessment of the platform's user base for AMR detection and data management.
4. Development of a multi-omics platform adept at integrating and analyzing standard microbiological data, alongside genomic data, including pathogen-specific sequencing data.

To effectively measure and evaluate these KPIs, we will implement a robust data-tracking dashboard. This dashboard will furnish us with data-driven insights into user enrollment, retention rates, and the geographic distribution of our users. Additionally, we will institute a continuous feedback collection mechanism from our users. This proactive approach will enable direct communication with users, facilitating prompt issue resolution and ongoing improvements.

To assess the effectiveness and robustness of our solution, we will need a consortium comprising various data-generating stakeholders from academia, research institutions, and hospitals. Managing these stakeholders' complexity can pose a significant barrier to demonstrating the utility of our platform in constructing a connected AMR database. Additionally, our lack of laboratory infrastructure impedes our ability to generate diverse AMR datasets for testing our solution's interoperability as a multi-omics platform.

We are currently implementing the "Decipher AMR project," a community wastewater AMR surveillance initiative aimed at independently generating AMR resistome data. This involves utilizing high-throughput methods such as quantitative polymerase chain reaction (qPCR) and whole metagenomic sequencing (WMG) with nanopore sequencing to train our AI/ML algorithm. This collaboration approach to data generation is helping us address the data inaccessibility issue that would otherwise hinder our ability to train our AI/ML algorithm model.

We are applying to the Trinity Challenge due to its specific focus on Antimicrobial Resistance (AMR). We believe this challenge offers an opportunity for our startup to access mentorship and financial resources, which have been elusive since our inception. Over the past 3 years, we have been diligently developing our solution, relying primarily on support from family and friends and small grant funding. However, the funding we received was insufficient to assemble a full-time team of developers, establish operational processes, and acquire the necessary technical infrastructure for a comprehensive solution.

Despite these challenges, we adopted a frugal approach and managed to create a minimum viable product with room for further development. With the resources available through the Trinity Challenge, we envision overcoming these barriers. Specifically, we aim to build a scalable solution to address AMR in Low and Middle-Income Countries (LMICs). The mentorship to be provided by the Trinity Challenge can guide us in refining our solution. At the same time, the financial support will enable us to expand our team, enhance our technical infrastructure, and accelerate the development of a robust AMR solution tailored to the needs of LMICs.

Our solution will be executed through a partnership consortium comprising two key entities: the academic/research institution, Kwame Nkrumah University of Science and Technology, and the Africa Health Innovation Center, a global health research and innovation-focused organization. Kwame Nkrumah University of Science and Technology will leverage its extensive experience in biomedical research and antimicrobial resistance studies, particularly through its esteemed medical school. This partnership will enable us to access and analyze previously untapped academic AMR data, which are currently fragmented and underutilized.

The Africa Health Innovation Center, renowned for its successful implementation of health research initiatives across five African countries, serves as a strategic collaborator for our project in Ghana. Moreover, their expertise and established networks position them to facilitate and expedite our market entry efforts into new African markets, with Nigeria being a primary focus following the successful implementation of our project in Ghana.

Through a WhatsApp community page