In this interview, Meredith Landry (Managing Editor, Citeline) speaks with Chris Forsdyke (Head of Laboratory Services, hVIVO), Martin Larke (Principal NGS Biostatistician, hVIVO), Francesca Manni (Territory Account Manager, Illumina, and Chelsea Pratt (Biopharma Segment Marketing Manager, Bio-Rad Laboratories) about how droplet digital PCR (ddPCR) and next-generation sequencing (NGS) can be combined for viral surveillance across antiviral drug development, from sensitive detection and quantification through to whole-genome sequencing and host-response profiling.
To start, what role does molecular biology play in hVIVO’s infectious disease trials, and what were you hoping to cover in this session?
Chris: At hVIVO, our molecular biology analytical strategy is critical in this space. We work closely with both Bio-Rad and Illumina, and we put their latest technology and software to use to reduce turnaround times, analytical costs, and sample volumes, while also expanding the data we can collect from highly valuable clinical trial samples.
The goal here isn’t just to talk about new lab platforms. It’s also to give a clear view of where these technologies fit in clinical development. No two projects are the same, and building the right molecular tools, along with the associated analytical and data workflow, matters whether you’re analyzing small or large clinical sample sets, with single or multiple targets of interest.
Standard qPCR has been the workhorse for decades. What does partitioning-based ddPCR offer that bulk amplification can’t?
Chelsea: The core difference comes down to partitioning. In traditional qPCR, you have a bulk reaction, typically 10 to 20 microliters, where fluorescence is read at the end of every amplification cycle. qPCR was our first real step at truly quantifying nucleic acids, but it requires near-perfect amplification efficiency to give accurate results, and it relies on standard curves.
In droplet digital PCR, you take that bulk reaction and partition it into nanoreactions, so the limiting factor becomes whether any individual droplet contains your target DNA or not. That enriches rare events so they can amplify without competing with highly abundant molecules. With qPCR, low-abundance targets can sit very close to your no-template control, which raises questions about validity. In ddPCR, your positive droplets are clearly separated from negative droplets, giving you confidence and precise calculations.
Quantification uses Poisson statistics: it’s the ratio of positive droplets to total droplets that gives you absolute quantification in copies per microliter, without a standard curve. And because the sample is divided across thousands of droplets, ddPCR can also be more tolerant of inhibitors than a single bulk qPCR reaction.
With the QX700, that entire workflow is contained in one instrument. It adds seven-color detection, enabling high-plex assays, and can support up to a 21-plex assay. It’s also automation-friendly and supports up to 768 samples per day.
From a cost point of view, ddPCR can be more cost-effective than traditional qPCR once you account for reaction volumes, replicates, and standard curves. A single qPCR target might use around 25 microliters of reagent per reaction, whereas a ddPCR reaction is around five microliters. With qPCR, you also typically need extra wells for standards to generate a curve, while ddPCR is absolute, so you can often reduce the total number of reactions needed.
You used Bordetella pertussis as a case study for pathogen detection. What makes it a useful example when evaluating a detection approach?
Martin: At hVIVO, we’ve been developing and using ddPCR across our research programs, and one of the key application areas is pathogen detection. We run ddPCR as part of an end-to-end workflow that includes sample extraction, reaction setup, droplet generation, and analysis. With automation in place, we can handle clinical samples in a consistent, controllable, and scalable way, which matters in infectious disease work where sensitivity, reproducibility, and turnaround time are all critical.
That workflow also isn’t pathogen-specific. We can apply the same automated process to a much broader set of targets, DNA or RNA, so it supports applications across discovery, translational research, and clinical development.
Whooping cough is a good illustration of the problem. It’s caused by Bordetella pertussis, and from a diagnostic perspective, culture can detect the bacterium, but it requires rapid transportation and can take weeks because the organism grows slowly. That makes it poorly suited to settings where you need a quick answer. qPCR is widely used because it’s faster and more sensitive, especially early in infection, but you’re often working with low-level clinical samples where assay design and controls really matter.
A key challenge is species differentiation. Many assays target high-copy insertion sequences (IS elements) for sensitivity, but those sequences can be shared across multiple Bordetella species, so you lose specificity. There are more specific targets, such as the pertussis toxin promoter (ptxA), but they’re low copy number, which hurts sensitivity. In practice, you’re balancing sensitivity against specificity, and multiplexed PCR approaches are often needed to bring those together.
Turning to sequencing, can you walk us through the NGS workflow from extracted RNA to actionable sequence data, and what the MiSeq i100 Plus adds?
Francesca: The current hVIVO setup for genomic profiling of viruses is based on an end-to-end workflow that supports the whole process from sample prep to analysis. It starts with extracted RNA that’s made sequencing-compatible using our library prep kit. That’s followed by the sequencing run, using Illumina’s sequencing-by-synthesis chemistry on the flow cell.
The raw images generated on the sequencer are converted into BCL files and then into FASTQ files. Those FASTQ files feed into secondary analysis, which aligns the sequenced genome to a selected reference genome and identifies differences. The library prep kit in use comes with its own analysis pipeline, the DRAGEN Targeted Microbial App, which matches sequencing data with reference genomes and identifies genome variations and single-nucleotide polymorphisms (SNPs).
The library prep kit currently in use is the Illumina Microbial Amplicon Prep (IMAP). This kit is based on PCR amplicons generated by the addition of PCR primers in the downstream protocol. Prepared libraries are then sequenced on the MiSeq i100 Plus, followed by analysis using DRAGEN genome alignment.
The MiSeq i100 Plus ships with onboard DRAGEN, so analysis can run on the instrument itself and supports multiple workflows locally. It also supports flow cells spanning from low throughput at 5 million reads up to 100 million reads. Advances in chemistry and imaging also support faster turnaround, with up to 30 gigabases in as little as eight hours, and it supports longer read configurations, including 2×500.
Meredith: What does hVIVO’s automated NGS infrastructure look like in practice, and why put so much emphasis on automation in library preparation?
Martin: At hVIVO, we’ve built NGS into an integrated process that starts with automated nucleic acid extraction. We use liquid handling robots to set up reaction plates and QIAGEN’s QIAsprint Connect machine to perform the extraction itself. Library preparation is also automated on a Hamilton Microlab STAR IVD, and that’s complemented with quality control using the Invitrogen Qubit and Agilent TapeStation before sequencing on the MiSeq i100 Plus.
Library prep can involve several stages of enzymatic reactions followed by cleanups and reagent additions, often with tight timing and low reaction volumes. Automating these steps helps us execute complex protocols precisely and reliably, especially when we’re working with large sample sets or multiple studies in parallel. It also helps minimize cumulative pipetting variation or sample handling errors that can show up later as sequencing variability.
We then use Qubit and TapeStation QC to quantitate and assess library quality before sequencing, which reduces rework and gives early visibility of issues. From a client perspective, the benefit is more predictable data generation and faster turnaround overall.
Turnaround time also depends on how well the workflow and data flow integrate with our LIMS, and our push toward paperless operations. Through a combination of process optimisation, laboratory automation, data integration, and more efficient reporting wherever possible, we are continuously trying to help minimise turnaround time for our clients.
Influenza sequencing is now routine in public health surveillance. How has that changed what regulators expect to see in antiviral development, and how do you interpret those sequences in a study setting?
Martin: Rather than relying on phenotypic or cell culture-based assays, viral research has increasingly become sequence-first. Sequencing lets you identify viruses and track them at the genomic level, often directly from clinical samples, which supports near-real-time monitoring of viral evolution.
Large agencies such as the CDC in the US sequence around 7,000 clinical influenza samples per year. They do that to determine how genetically similar circulating flu viruses are, track evolution, and detect changes that may influence transmissibility, severity, or antiviral resistance. They also use this data to support strain selection for future vaccines. Last year, for example, a new H3N2 influenza subclade (subclade K) rapidly rose to become the dominant circulating virus worldwide, and that kind of shift is captured through systematic sequencing.
In drug development, this is showing up in regulatory expectations. When you submit an antiviral drug package, there’s an increasing requirement for NGS data to investigate whether resistance mutations emerge during treatment. Regulators expect any link between a sequence change, drug exposure, and clinical outcome to be clearly explained.
A key part is choosing the right reference context. At hVIVO, we maintain a database of season-matched circulating influenza strains built from recent whole-genome sequencing datasets. Comparing only to older historical references can show large numbers of differences that reflect normal background evolution, rather than changes that emerged under antiviral pressure during a study. Using season-matched references helps us focus on variants that emerge during the study and are more likely to be relevant for resistance assessment.
The end result is a regulatory-ready output, such as a variant frequency table, showing sequence changes, their prevalence, and how they track over time in relation to treatment. We also verify we have sufficient sequencing depth across regions of interest and regulatory importance, including the hemagglutinin gene.
Beyond pathogen genomes, what does transcriptomic profiling via RNA-Seq add compared with measuring one blood biomarker at a time?
Martin: This shifts the focus from single biomarkers to more of a systems-level view of a drug response. Traditionally, blood-based biomarkers such as proteins, cytokines, chemokines, or soluble receptors are measured one analyte at a time across a large sample set. Those measurements are valuable, but they can be limited in the biological context they provide.
For example, you might see a decrease in a single analyte like interleukin-6, which suggests a drug effect, but it’s hard to tell if that’s isolated or part of a broader inflammatory program. With gene expression profiling, you can look across the transcriptome. By using three-prime RNA-Seq, sometimes referred to as QuantSeq, we can measure gene expression across essentially the entire human transcriptome, using RNA as a readout of gene activity.
At the pathway level, instead of examining one gene in isolation, we can assess coordinated changes across biological programs. Gene set enrichment analysis can show, for instance, increased apoptotic signaling alongside reduced inflammatory responses. Tools like volcano plots then summarize the gene expression changes following treatment, making it easier to see which genes change the most and which changes are more statistically convincing. In drug development, this helps with mechanistic understanding, target and biomarker selection, and characterizing variability in response during clinical trials.
If you look across ddPCR and NGS together, what does combining these platforms with hVIVO’s controlled human challenge models enable that neither could achieve on its own?
Martin: We’re focused on building advanced molecular biology solutions into automated, scalable workflows that generate decision-ready data. With digital PCR and NGS, supported by automated extraction, assay setup, and library preparation, we can develop assays targeting DNA or RNA sequences across a broad range of targets, including bacterial pathogens, viral genomes, and the host transcriptome.
These capabilities are complementary rather than siloed. ddPCR can provide highly sensitive, quantitative answers for defined targets, while NGS can add genome-wide resolution for surveillance, resistance analysis, and deeper characterization.
One of the key differentiators for hVIVO is combining those technologies with our clinical expertise, particularly through controlled human challenge models. That lets us interpret molecular and genomic data within a well-defined infection timeline, alongside treatment exposure and clinical endpoints, which is important for translational insight and for regulatory confidence.
Can you use IMAP for other viruses beyond influenza?
Martin: Yes, you can. The IMAP procedure is very flexible. The example we showed was targeting influenza, so the first part of that process uses primers to map to the influenza genome, but those can be swapped out with other primers. One of the applications we’re developing is RSV whole-genome sequencing, and we plan to expand that across other viruses.
With our automation workflow as well, it’s likely we’ll move into broader monitoring panels. There are viral surveillance panels available that can detect the presence of up to around 200 different viruses of public health concern, and we’re investigating how to build workflows around that, so we can generate whole-genome sequencing data from a wider range of viruses.
Antivirals have been hard to develop. How do you expect more sequencing information to affect novel antiviral therapeutics, and do you see a shift in modality?
Chris: In the infectious disease space we operate in, we’re seeing a resurgence in antiviral projects. Those antivirals range from small molecules through to antibodies and monoclonal antibodies. Using these platforms in parallel as part of an overall analytical strategy is important in this space.
There are guidance papers that require the use of NGS for variant analysis in small-molecule antivirals, to understand what’s happening with these newer therapeutics. The application continues to expand, and the projects we’re putting this technology to use for are diversifying as our analytical pipeline moves from the traditional vaccine space toward antivirals, including newer types of antivirals.
About Chris Forsdyke, Martin Larke, Francesca Manni and Chelsea Pratt
Chris Forsdyke is Head of Laboratory Services at hVIVO. He brings 25 years of strategic and operational leadership across global drug development and manufacturing enterprises, with experience spanning the UK, EU, and US. His focus is on building business and operational frameworks and leading high-performance teams.
Martin Larke is Principal NGS Biostatistician at hVIVO. He works on assay design and statistical methodology for next-generation sequencing, supporting = infectious disease research. He contributes to hVIVO’s biostatistics and methodology services across the drug development lifecycle.
Francesca Manni is Territory Account Manager 2 at Illumina. She previously worked as a senior inside sales account manager and senior scientist at the company, and her background includes research in organic chemistry, optoelectronics, imaging, and sequencing.
Chelsea Pratt is the Biopharma Segment Marketing Manager at Bio-Rad Laboratories. Dr Pratt holds a PhD in Biological Chemistry from the University of Texas Southwestern Medical Center.
About hVIVO
hVIVO plc is a science‑led early‑phase drug development company purpose‑built to meet the growing complexity of modern clinical research. The Company operates an integrated early‑phase ecosystem that combines specialist clinical sites, advanced virology and immunology laboratories, human challenge expertise, and early drug development consulting. This unified model enables sponsors to generate rigorous, decision‑ready human data earlier in development, reducing uncertainty and accelerating progression through Phase I and II trials.
With industry‑leading capabilities in respiratory and infectious disease, alongside expanding expertise in cardiometabolic and other high‑growth therapeutic areas, hVIVO supports a diverse global client base that includes seven of the world’s ten largest biopharmaceutical companies. Its London quarantine facilities are the largest purpose‑built human challenge units in the world, complemented by additional early‑phase clinical capacity in Germany and a specialist consulting team providing strategic, regulatory, and biometry expertise.
The Company’s integrated approach delivers a seamless pathway from preclinical planning through early proof‑of‑concept, supported by continuous patient recruitment through FluCamp and a network of outpatient clinical sites for Phase II and III studies. By unifying scientific insight, operational control, and advanced laboratory capabilities, hVIVO provides sponsors with the clarity, speed, and reliability required to advance new medicines with confidence.