MEDICUS: Could you explain pathogen genomics a little more simply?

Barber: Pathogen genomics is the use of a pathogen’s genetic information to understand its behavior. That includes whole genome sequencing, which gives a very complete picture but can be costly, and targeted next-generation sequencing (tNGS), which is kind of more zoomed-in genomics that may be especially relevant in studying antimicrobial resistance genes or spike proteins involved in infection.

Suci: And AI helps us to contextualise these genomic insights with other forms of public health data. That might include demographic, socioeconomic or mobility data, depending on the use case. Bringing these together could reveal more patterns that might otherwise be missed.

MEDICUS: What are the biggest challenges in the field right now?

Suci: Data-sharing is one of them. Since COVID, many countries have increased their genomic-sequencing capacity, which means more data is being generated. But not all of that data is shared, analysed further or used optimally. Concerns about data sovereignty remain a major obstacle to improving regional health security.

That matters because infectious diseases respect no borders. One country’s preparedness affects another’s. Indonesia's capacity to manage an outbreak, for example, also benefits Singapore. No country is alone here.

MEDICUS: Can you explain what data sovereignty means in this context?

Suci: In this region, these concerns are very real and also shaped by history. One example: in 2007, Indonesia refused to share H5N1 samples due to concerns that sharing samples would only benefit higher-income countries. The country providing the samples—Indonesia— hit a paywall to get fair access to medical countermeasures produced from the sample. Such historical precedents have influenced how countries think about sharing raw, sensitive data.

Barber: Yes, I think it is the question of control and benefit. If a country owns the data, it can decide who has access to it and how it is used. What often drives concern is the feeling that academic, commercial or R&D benefits are distributed unevenly.

MEDICUS: How does PathGen address that?

Suci: The sovereign-by-design approach is central. Countries need to be able to retain full authority over ownership and how intelligence is shared if we want the platform to be trusted and genuinely useful.

Barber: That trust is critical. You cannot build a regional system without acknowledging that these concerns are valid.

MEDICUS: PathGen involves partners and stakeholders across sectors. What does that collaboration look like in practice?

Barber: I think it's unique. It's quite beneficial to have different stakeholders, different perspectives. We have AI and biotech partners from the private sectors, alongside university and public sector partners. I think it's one of the strengths of PathGen.

Suci: We have clinicians, public health researchers and tech innovators all trying to tackle the same problem. Duke-NUS, SDGHI, the University of Sydney, IXO, Sequentia Biotech and Amazon Web Services all bring different perspectives. And Asia PGI has also played an important anchoring role in this regional work throughout.

MEDICUS: How do you make that kind of collaboration work well?

Suci: Open communication and shared vision. In the core team, we maintain regular communications. Even though we may speak different languages, or use different terminologies, as much as possible, we are aware that we are working together in an environment that requires multidisciplinary collaboration to tackle a problem this big. So that's what we've been doing. I think I’ve learnt a lot from Mike about certain terminologies and perspectives.

Barber: We organise much of the work in cross-functional squads that operate in short sprints. Borrowing from the startup playbook, this structure enables teams to focus intensely, review what we've done, align with the partners, and then repeat. This is how we have managed to keep the different partners working fluidly together.

Having so many different perspectives, you do learn each other's languages a little bit. So I think having Suci's, Duke-NUS' and Asia PGI's input into the public health side of things has been fantastic and incredibly unique.

MEDICUS: What advice do you have for students who may be interested in getting into this sector or field of study?

Barber: The multidisciplinary angle is key, especially with AI, because AI enables a lot of work to get done that previously required specialist skill sets. There's going to be more benefit in being multidisciplinary and working across different disciplines, languages and technologies. For students preparing for AI in the workplace, it's a tough one just because even in the last six months, AI has made leaps and bounds. Tech has always been a very fast-moving field, but I think the velocity of that change has increased significantly.

So I suspect any advice I give would probably be out-of-date by the time this goes out. The broader advice I could give would be just to stay up-to-date on, for instance, cutting-edge things. It's good to have a little bit of technical understanding of what AI actually does, just so you can avoid some of the pitfalls. But otherwise keeping up-to-date and staying on the leading edge of AI will be in itself a massive advantage.

MEDICUS: How will these rapid advances influence the way forward for Pathgen, since it uses AI?

Barber: Pathgen is still on the cutting-edge of AI. A lot of what we're doing is at the forefront of what is capable with today's models. I think for PathGen, continuing to utilise frontier capabilities is going to be very important. I can give you a very specific example, which is this idea of skills. Different AI agents can have dedicated skills to different jobs. So you could have a skill that does data organisation, one that makes PowerPoints, one that interrogates genetic data. And so onboarding these skills onto the AI, essentially, enables you to grow your AI in a modular fashion, which is going to be useful for iterating.

MEDICUS: What are you most excited about working on in PathGen, and what are you most looking forward to?

Barber: I think for me it's the impact, it's the data sharing. Like I said, some of these data sets never see the light of day. The more genomic data we can surface, the more prepared we can be for eventualities. I think it’s really one of the most exciting promises of PathGen.

Suci: For me, it's how the platforms get deployed to empower local health systems, and to see how the platform is installed on the ground and facilitates decision-making locally, while complying with national standards for data security.

MEDICUS: Is there a broad roadmap that you could describe for Pathgen? And what is your dream for PathGen?

Barber: We're currently building the technical MVP (minimum viable product), so we're doing that with input from the country partners. It is an iterative process of improvement, with feedback from potential users. We're working on what might be thought of as a pre-alpha version at the moment.

Suci: We are continuously working with country partners to scope out their priority use cases on how to operationalise PathGen. For now, the platform is based on use cases for TB (tuberculosis), and then we want to develop it further for other use cases, as well. For example, maybe arboviral diseases, wastewater diseases, antimicrobial resistance, and so on. That's what we are currently working on with our country partners. And then, of course, to pilot the platform on the ground and invite users to give feedback for cross-correction.

Ultimately, the roadmap is focused on capacity building. It's about training national public health labs and ministries to develop their own capacity to use the data and transform that into actionable insights. So that's something that we are also working on through our partnerships.

Barber: For me, it's back to the data. I think we have multiple countries that have the intelligence that would benefit the region. So my dream is for these sequences that have not been shared in many years to be shared.