Author: admin

Posted on

AI Could Help More Donor Hearts Reach Transplant Patients

Integrating artificial intelligence (AI) tools into transplant infrastructure could save a significant amount of available donor hearts from being discarded, according to research presented at the International Society for Heart and Lung Transplantation (ISHLT) 46th Annual Meeting and Scientific Sessions.

“There is a massive shortage of heart donors in the United States, with patients waiting months—if not longer—for a transplant, often on life support in the ICU. So the stakes are very high,” said Brian Wayda, MD, transplant cardiologist and assistant professor of medicine at NYU Grossman School of Medicine. 

Despite an ongoing shortage of donor hearts, only up to 40% of the hearts that become available are actually transplanted. Transplant teams will typically evaluate potential donors based on a series of donor risk factors, including the person’s age, disease history, and drug use record, among others. However, evidence is still limited on how each factor affects post-transplant outcomes, and decisions need to be made quickly to ensure any suitable hearts find a matching recipient on time. 

“It’s an extremely complex judgment call that must be made in a very short time window, often in the middle of the night,” said Wayda. “AI can support these life‑and‑death decisions made under extreme time constraints.”

Together with scientists at Stanford and other leading U.S. research centers, Wayda has developed a web-based prediction tool called TOPHAT (Tool Predicting Heart Acceptance for Transplant). This machine learning algorithm evaluates 20 donor characteristics to estimate how likely a transplant center is to accept a donor heart, based on historical data from over 78,000 potential donors.

Using this tool could help experts make decisions in a more data-driven, consistent, and efficient way. This could reduce the likelihood that a suitable donor heart gets discarded due to time running out before a matching recipient is found. 

“The tool doesn’t say ‘this is a good heart’ or ‘this is a bad heart,’” Wayda explained. “Instead, it quickly shows how a donor compares to the national experience. An older donor, or one with a single risk factor like cocaine use, may look high-risk at first glance. But when you consider all the variables at once, that donor may not be any riskier than a typical heart we already use.” 

There are currently over 4,000 patients waiting for a heart transplant in the United States. Even a relative increase of 500 additional hearts becoming available each year would be enough to reduce wait time substantially, said Wayda.

Going forward, the researchers are working toward developing a unified decision support system that brings together output from TOPHAT and other AI tools, as well as the broader donor medical record, to generate a single, easy-to-digest summary for clinicians making time-sensitive decisions about a potential transplant. 

“The real value of AI is helping us synthesize a huge amount of data quickly and objectively so clinicians can make better-informed choices,” said Wayda. “With this kind of integrated view, doctors would be less likely to anchor their decision on a single ‘red flag’—such as donor age over 50—and decline hearts that could have performed well.”

The post AI Could Help More Donor Hearts Reach Transplant Patients appeared first on Inside Precision Medicine.

Posted on

Gut Microbiome Signatures Predict Melanoma Response to ICB Treatments

Researchers at NYU Langone Health’s Perlmutter Cancer Center have found that patterns in the populations of bacteria in the gut microbiome can predict which melanoma patients are more likely to benefit from immunotherapy. The study, published in Cell, showed that specific bacterial signatures, when analyzed in the context of a patient’s overall microbiome profile, can forecast cancer recurrence after immune checkpoint blockade (ICB) with accuracy as high as 94%. The findings suggest that using this information could help identify which patients will respond to ICB treatment and which are more likely to relapse.

“Our study identified for the first time gut bacterial types that can serve as markers of increased recurrence risk in these specific patients, which will help to tailor treatment,” said study senior author Jiyoung Ahn, PhD, a professor of population health at NYU Grossman School of Medicine and associate director of population research at NYU Langone’s Perlmutter Cancer Center.

ICB is a form of cancer treatment that enhances the immune system’s ability to recognize and attack tumor cells. Drugs such as nivolumab and ipilimumab work by inhibiting molecular “checkpoints” that normally restrain T cell activity to allow immune cells to mount an anti-tumor response. Because of the success of ICBs in advanced cancer, this form of treatment is now expanding into earlier-stage, higher-risk patients following surgery.

“Immune checkpoint blockade (ICB) therapy has transformed the management of advanced, unresectable melanoma,” the researchers wrote. However, it is not effective for all patients. “Clinical benefit remains unpredictable, with approximately 25%–40% of patients experiencing disease recurrence despite therapy,” they added.

In their search for biomarkers that could stratify responders from non-responders, the NYU investigators analyzed stool samples from 674 melanoma patients enrolled in the Phase III CheckMate 915 clinical trial. Participants had undergone surgical tumor removal and then received either a combination of nivolumab plus ipilimumab or nivolumab alone for up to one year. Using shotgun metagenomic sequencing, the researchers characterized the gut microbiome at strain-level resolution before treatment and, in a subset of the patients, during therapy.

Their analysis identified bacterial taxa, including Eubacterium, Ruminococcus, Firmicutes, and Clostridium, that were associated with recurrence risk.

An important finding was that predictive accuracy was dependent on matching patients by their overall microbiome composition. “Recurrence prediction was strongest when the validation cohort exhibited GMB profiles similar to those in the discovery cohort,” the researchers wrote. When patients were closely matched based on microbial similarity, prediction performance reached area under the curve (AUC) values between 0.78 and 0.94. “This evidence indicates that taxonomic markers for prediction of recurrence are generalizable across regions for individuals with similar GMB composition,” the researchers noted.

The study’s design sought to address a longstanding challenge in microbiome research, notably that earlier studies had shown bacterial markers linked to immunotherapy response varied widely by geography.

“Past studies have struggled because the gut bacteria that predict treatment success seemed to change from one region to another,” Ahn said. “Our study provides a new method that overcomes this barrier, showing that these markers are indeed generalizable if we account for the person’s underlying microbiome.”

The study also showed that the gut microbiome remains stable during treatment, a finding that suggests the potential to manipulate the gut microbiome before therapy begins. “This stability suggests an important window of opportunity before treatment begins,” Ahn told Inside Precision Medicine. “We are currently planning diet-based intervention trials aimed at actively modifying the microbiome prior to immunotherapy. The goal is to move beyond observational associations toward actionable strategies that can improve treatment response.”

The biological mechanisms underlying these associations may relate to how gut bacteria influence immune activity. “These taxa are largely fiber-metabolizing bacteria that produce short-chain fatty acids, such as butyrate,” Ahn said. “These metabolites are known to play important roles in modulating immune function, including enhancing anti-tumor immune responses and regulating inflammation.” The researchers also noted links between these bacteria and metabolic pathways such as “glycolysis/gluconeogenesis” and the “pentose phosphate pathway,” which prior research has shown can affect cancer treatment outcomes.

Evidence supporting the microbiome’s role in immunotherapy response has been accumulating. Prior studies in metastatic melanoma have shown that fecal microbiota transplantation can restore responsiveness to ICB in some patients, via activation of CD8+ T cells. But earlier research has been limited by small sample sizes and regional variability.

The current study, however, examines the influence of the microbiome during adjuvant therapy and provides a potential method for overcoming geographic differences.

“The main challenge is that prediction models may be limited to subsets of populations with similar underlying microbiome structures,” Ahn noted. “Moving forward, we will need well-characterized, large-scale microbiome reference datasets that allow appropriate matching across populations and regions.”

Additional work is needed in order to use these signatures in the clinic. “The next steps include validation in independent cohorts and prospective trials,” Ahn noted. “Ultimately, these biomarkers have the potential to guide patient stratification and optimize immunotherapy outcomes in clinical settings.”

The post Gut Microbiome Signatures Predict Melanoma Response to ICB Treatments appeared first on Inside Precision Medicine.

Posted on

Hearing Loss Gene Therapy Lasts More than Two Years

A trial of a gene therapy to treat people with hearing loss related to recessive mutations in the OTOF gene shows the treatment is effective and safe for at least 2.5 years.

The study, published in Nature, showed around 90% of those who received the adeno-associated viral (AAV) vector gene therapy showed at least some restoration in hearing.

Improvement was rapid in the first six weeks, improved further by 26 weeks and in a small subset of patients remained stable for 2.5 years of follow-up.

“It’s remarkable to see patients go from complete deafness to being able to hear,” said the study’s co-lead author, Zheng-Yi Chen, PhD, the Ines and Fredrick Yeatts Chair in Otolaryngology and an associate scientist at Massachusetts Eye and Ear hospital, in a press statement. “For many patients, that also means the ability to develop and use speech.”

The OTOF gene encodes the otoferlin protein, which is critical for normal hearing. When otoferlin is missing or nonfunctional, inner‑ear hair cells can’t relay sound information to the brain, leading to severe or complete deafness. This kind of hearing loss is rare and inherited in a recessive manner, needing mutations from both parents for a child to be affected.

As of this year there are at least five gene therapies being developed to treat this kind of deafness, for example, by Akouos/Eli Lilly and Decibel/Regeneron in the U.S., Sensorion in France, and at least two additional programs in China.

The current study took place in China and included 42 people between the age of eight months and 32 years (average age six years) and is the largest cohort of OTOF gene‑therapy patients reported so far, as well as the longest study follow-up period.

The participants received one of three doses of the AAV gene therapy injected into their cochlea’s and were followed up for 13 weeks to 2.5 years (median 52 weeks) to assess the impact of the therapy on hearing and also to evaluate safety.

Overall no serious adverse events or dose-limiting toxicities occurred. Around 90% of participants experienced hearing restoration to some degree with fast improvements seen in the first six weeks after treatment and slower improvements after that. A subset of patients (seven ears from seven patients) were included in the 2.5 year follow-up group and results were similar to those seen at two years.

Some groups did better than others. For example, hearing restoration was 100% in children aged up to three years and 92% in those aged 3-8 years. Improvement was seen in older children and adults, but to a lesser degree than that seen in young children in the study. Participants with better outer hair cell function on enrollment also responded better to the therapy than those with greater functional loss.

“It is very encouraging to see meaningful improvements in some adult patients. It suggests there may be more flexibility in the human auditory system than we expected,” said Chen, who is also the scientific founder of Salubritas Therapeutics, a Massachusetts based biotech focusing on hearing loss correction.

The post Hearing Loss Gene Therapy Lasts More than Two Years appeared first on Inside Precision Medicine.

Posted on

Viral Contamination Still a Challenge for CGT Industry

Raw material testing will remain the foundation of cell and gene therapy (CGT) sector quality control strategies for the foreseeable future, according to new analysis, which shows the industry still lacks suitable virus detection and inactivation methods.

Biopharmaceutical raw materials—the culture media ingredients, the reagents, and even the production cell lines themselves—are the biggest source of viral contamination in drug manufacturing.

To mitigate the risks, the protein drug industry has developed downstream virus detection, inactivation, and removal strategies to make sure products do not pose an infection risk.

For CGT firms, ensuring products are virus safe is more of a challenge, says Yoshiaki Maruyama, PhD, from the office of cellular and tissue-based products at Japan’s Pharmaceuticals and Medical Devices Agency (PMDA).

“Viral contamination of CGT products may arise from virus-contaminated raw materials or ancillary materials of human or animal origin or from the inadvertent introduction of viruses during the manufacturing process.

“Appropriate raw material controls and robust quality control parameters must be established and maintained throughout the manufacturing process to effectively manage the risk of viral contamination,” he tells GEN.

Inactivation and removal challenges

The big problem is that cell and gene therapies are too sensitive to survive current viral inactivation methods, most of which were developed with protein therapeutics in mind.

Maruyama says, “Most conventional virus inactivation or removal processes inevitably result in cell damage or loss in cell therapy and tissue-engineered products or adversely affect viral vectors in gene therapy products.”

As a result, CGT sector quality control efforts have focused on screening raw materials and finished products, according to Maruyama, who looked at current regulations and common approaches in a recent study.

“In the CGT sector, viral safety is achieved by implementing a comprehensive viral testing program. The use of inactivation and removal processes is challenging for CGT products and raw materials, so quality control strategies relying on screening are generally used,” he says.

Technological solutions?

In future, technologies may play a greater role, according to Maruyama, who says, “

“NGS technologies are expected to be applicable to the detection of adventitious viruses in human or animal cells. NGS offers a powerful, unbiased approach for detecting known and unknown viral contaminants,” they write.

However, as the authors point out, further development will be required as NGS systems detect nucleic acids rather than viable, infectious virus particles.

“Currently, there are no globally accepted NGS-based procedures or validated analytical methods that have reached a consensus on their use as substitutes for conventional viral tests. Therefore, the use of NGS as an alternative to conventional viral tests, including reducing the use of experimental animals, requires further evaluation depending on the specific test to be replaced,” they write.

And in the future, artificial intelligence (AI) systems may also play a role.

“This is largely speculative, and there are currently no concrete examples, but AI-based tools have been applied to manufacturing control for deviation prediction and similar approaches might also be useful for controlling viral contamination risks in CGT products and raw materials,” he says.

The post Viral Contamination Still a Challenge for CGT Industry appeared first on GEN – Genetic Engineering and Biotechnology News.

Posted on

AI Wizard Adapts Processes in a Self-Driving Lab

German researchers who run a self-driving laboratory have created an agentic AI wizard to help their students rapidly design and implement new processes.

The wizard, which uses N8N software, can guide a student through establishing experiments without the need for coding, allowing them to quickly set up a new process.

According to Matthias Franzreb, PhD, a professor and departmental leader in bioengineering and biosystems at the Karlsruhe Institute of Technology, developing wizards could help any autonomous laboratory where the experimental setup needs to change fast.

“Each of our bachelor’s and master’s students has their own type of experiment and, in the beginning, going into Python scripting, it used to take two months to have the whole thing programmed,” he says.

By contrast, he says, the AI agent can help the student develop a new process within one or two days. It has so far been used to develop around six processes, he says, for a slightly larger number of students, as the same template can be used more than once.

Bioprocessing, like many other areas of human endeavor, is experiencing disruptive change with the growing use of digital tools at both the laboratory and commercial scale, Franzreb explained in a talk at Bioprocessing Summit Europe.

Among these changes is the difference between classical labs, which have automated equipment, such as liquid handling stations, but where scientists must design and set up their own experiments, and self-driving labs. In the latter, he explains, machine learning uses a first set of experiments to autonomously decide what experiments should be next.

In his talk, Franzreb also showed how a wizard could be used for designing a chromatography experiment. An experiment was set up to determine batch parameters at a small-scale in 96-well plates. From this, the software used a chromatography simulation to find the optimal conditions for the experiment and then ran it in a real chromatography system to validate the results.

According to Franzreb, the next step for the self-driving laboratory will be working with the German Research Center for Artificial Intelligence (DFKI) and other research partners to develop ontological capabilities for the wizards so they can extract context for the experiments from Standard Operating Procedures (SOPs) or the academic literature.

“I think this is simple in principle,” he explains. “But at the moment we don’t have it, and it will be a challenge to roll out.”

The post AI Wizard Adapts Processes in a Self-Driving Lab appeared first on GEN – Genetic Engineering and Biotechnology News.

Posted on

Monitoring Mammalian and Microbial Bioprocesses in Real Time

At the 2026 BiOS conference in San Francisco, researchers presented a biosensing platform aimed at improving how living cells and tissues are monitored during drug bioprocessing. Known as TissueSense, the system provides continuous, real-time insight into cellular behavior without disrupting the biological environment.

In biopharmaceutical manufacturing, maintaining consistent cell health and productivity is essential. Yet many monitoring approaches still rely on intermittent sampling or endpoint measurements, offering only partial visibility into dynamic biological processes. TissueSense addresses this limitation by enabling continuous, in situ observation—capturing changes as they unfold.

The platform combines resonator-based photonic sensing with phase contrast microscopy, allowing simultaneous detection of biochemical activity and structural changes in cells. This dual approach provides a more complete picture of how cells respond to process conditions, such as nutrient shifts or environmental stress, which directly impact production outcomes.

A defining feature of the system is its label-free operation. Conventional biosensing methods often require fluorescent markers or reagents that might alter cell behavior or limit long-term monitoring. By removing these constraints, TissueSense supports extended observation of living systems in conditions closer to their natural state, an advantage for prolonged bioprocesses.

Data from the platform are analyzed using machine learning to simultaneously quantify up to 18 biomarkers, linking molecular outputs—such as secreted proteins—to tissue structure and function. This multiplexed capability is particularly relevant in drug manufacturing, where small variations in cellular activity can influence yield, quality, and reproducibility.

While TissueSense focuses on mammalian tissue models, parallel advances in microbial systems highlight a broader shift toward continuous, high-resolution monitoring across bioprocessing platforms. In yeast-based systems, for example, researchers have developed microbead-based cultivation methods that enable high-throughput, label-free screening of millions of individual mutants in extremely small volumes. These approaches can enrich desirable traits, such as resistance to metabolic inhibitors, by thousands-fold, supporting strain optimization for industrial bioproduction.

Similarly, in bacterial bioreactors, automated flow cytometry techniques now allow real-time tracking of population dynamics and physiological states. By combining DNA staining with indicators of active replication, these systems provide continuous insight into growth rates and cell cycle behavior, helping optimize feed strategies and overall process performance.

Together, these developments point toward a more integrated future for bioprocess monitoring—one that spans mammalian, yeast, and bacterial systems. Continuous, non-destructive sensing technologies are enabling researchers and manufacturers to move beyond static measurements toward dynamic control of biological production.

The post Monitoring Mammalian and Microbial Bioprocesses in Real Time appeared first on GEN – Genetic Engineering and Biotechnology News.

Posted on

Overcoming the VLP Purification Bottleneck

Virus-like particles (VLPs) are a popular platform for biomanufacturers because of their good biosafety profile, immunogenicity, and ease of engineering, although downstream purification remains bottlenecked.

“Successful purification of VLPs cannot rely on any single unit operation, but instead requires integrated, product-specific process design guided by the critical quality attributes of the target particle,” Jingchao Zhang, PhD, Chengdu University of Technology, and Chen Chen, Tianjin University, point out in a recent review. The complex processes needed to generate VLPs result in multiple routes to success, and they are each vulnerable to environmental and process-induced stress, they note.

One option to mitigate such stress is buffer optimization. When developing buffers, they advise evaluating pH, ionic strength, ion species, excipients “such as nonionic surfactants,” and stabilizers that “improve thermal and freeze-thaw robustness.” Start by identifying conditions that most often cause the target molecule to fail, they advise. “This stress-informed characterization is particularly valuable because the stability of VLPs cannot usually be inferred from a single condition alone and may depend on both particle type and solution context,” they write.

Another option is “gentle chromatography.” By that, Zhang and Chen mean macroporous (100 nm or greater pore sizes) chromatography media that support process scaleup by improving binding capacity, increasing mass transfer rates and recovery, and are gentler on VLPs than the narrow (less than 30 nm) agarose media that often are used. Emerging options include non-woven structures and medium-to-large pore hydrogel microspheres, both of which have achieved success, respectively, with adeno-associated viruses and exosomes. “Overall, the chromatographic strategy for VLP purification should be framed as a balance between separation performance and particle preservation,” they conclude.

Process analytical technology is also increasingly valuable as technologies emerge to enable real-time monitoring, analysis, and control, they add.

Managing product heterogeneity is another challenge, as VLP downstream purification must address process- and product-related impurities. Typically, this is a multi-step endeavor “including clarification, ultrafiltration/diafiltration, chromatography, and, where appropriate, disassembly/reassembly-based purification,” Zhang and Chen report.

So far, there haven’t been many viral clearance studies for VLPs, they say. Part of the challenge is the many different expression systems used, such as E. coli, Chinese hamster ovary cells, and insect baculovirus expression vector systems.

In comparing the major downstream viral clearance strategies mentioned in the literature, Zhang and Chen report:

  • Solvent/Detergent (TritonX-100) treatment has some environmental concerns and mainly inactivates enveloped viruses.
  • Anion-exchange chromatography shows robust viral clearance only if the isoelectric points of the virus and VLPs differ.
  • Ion-exchange chromatography is constrained in high-salt or complex sample matrices.
  • Cation-exchange chromatography is highly effective in specific conditions.
  • Virus filtration is gentle and clears enveloped and non-enveloped viruses, but large VLPs may be larger than the filter pore size.

Looking forward, Zhang and Chen predict near-term VLP purification advances will include: responsive materials and media that enable precise control; AI and machine learning that predicts structure-performance relationships to accelerate materials screening; greater process intelligence; continuous processing; increased use of quality-by-design principles; parallel development in regulatory science; and clearer regulatory standards.

The post Overcoming the VLP Purification Bottleneck appeared first on GEN – Genetic Engineering and Biotechnology News.

Posted on
<![CDATA[Learn how antipsychotics and stress raise prolactin, what symptoms to spot, and when to test—plus practical options to lower levels.]]>
Posted on

Jurgi Camblong: Data-Driven Doctors Without Borders

Jonathan D. Grinstein, PhD, North American Editor of Inside Precision Medicine, hosts a new series called Behind the Breakthroughs that features the people shaping the future of medicine. With each episode, Jonathan gives listeners access to his guests’ motivational tales and visions for this emerging, game-changing field.

Precision medicine is often framed as imminent: gather more data, refine analytics, and individualized care will naturally follow. In reality, progress has been uneven. Genomic, imaging, pathology, and clinical data remain fragmented across systems and poorly integrated into clinical workflows. The core challenge is not data scarcity but the ability to interpret complex, heterogeneous inputs quickly enough to guide real medical decisions. To address this, Jurgi Camblong founded SOPHiA Genetics with a focus on building infrastructure rather than isolated tools—aiming to turn multimodal health data into actionable insights, a goal far more difficult in practice than in theory.

In Behind the Breakthroughs, Camblong highlights persistent structural and technical barriers limiting data-driven healthcare. Genomic standardization, for example, remains inconsistent, with approaches ranging from targeted panels to whole-genome sequencing, each balancing cost, sensitivity, and speed. The field is also shifting from single mutations to complex interactions among variants. Expanding beyond genomics adds further complexity, as transcriptomics, radiology, liquid biopsy, and computational pathology each involve distinct methods and clinical uses. Rather than enforcing uniformity, SOPHiA Genetics works across this diversity to produce consistent, clinically usable outputs despite technological and regulatory variation.

Ultimately, success depends on integrating statistical, machine learning, and deep learning methods while staying grounded in biology. A major limitation is the lack of robust feedback loops: precision medicine requires long-term patient outcomes, which many systems fail to capture. Without this, even advanced models are constrained. The central challenge is execution—translating existing data into meaningful insights that improve individual patient care.

This interview has been edited for length and clarity.

 

IPM: What types of multi-omics datasets are currently workable and applicable in a clinical setting, and how do you see their role evolving in routine patient care?

Camblong: When we started in 2015 and launched the platform into the market, people were just analyzing CFTR for cystic fibrosis and BRCA1 and BRCA2, two genes for hereditary cancer. To be honest, there were some efforts around whole genome analysis, but it was very, very rare. Our intent was always not to be a research tool but a tool that brings real benefit to most patients routinely and safely, and things evolved over time.

Now, probably the mean number of genes analyzed when producing genomic information for a patient is around 100 genes. Then you have some solutions that require analyzing only 30 genes because you want to be extremely precise, cost-effective, and rapid. There are other solutions that require sequencing the whole genome. But getting full information with the same sensitivity you can have with smaller panels is not an easy task, and this is where algorithms are really important.

In our case, the fact that we have grown along this journey with the field gives us an advantage today, enabling people to produce more genomic information with the same sensitivity as smaller panels. Genomics is continuously evolving. In the past, people did not necessarily look at copy number variations. Now we are even talking about partial copy variations, like in a gene called PTEN, which is a driver gene, and where a partial CNV can be very important.

What I am trying to explain is that it is not yet simple. It is not streamlined. Lab protocols are different; sequencing approaches are different; it is a constant evolution. In our case, being an operating system that supports thousands of hospitals, we are privileged to be exposed to this complexity, which enables us to improve our algorithms more rapidly and deliver them back to users who can benefit from new capabilities.

Transcriptomics is becoming a very interesting data modality. Initially, it was used to detect so-called gene fusions, specific genomic features that are hard to detect from DNA and require RNA. I am quite bullish on transcriptomics. I believe it will enable cancer subtyping at scale, possibly with more efficient methodologies than what is done today on tissue. It may not replace tissue, but it may allow us to go further and, in some cases, provide more objective outcomes than staining protocols.

Along those lines, radiomics is also very important. By radiomics, I mean data produced by radiologists, CT scans, PET scans, and MRI. There is a signal in this data. For example, you can see if cells are necrotic. You get additional information based on tissue composition and imaging. You can automatically measure tumor volume.

In metastatic cases, where tumors are spread, measuring them is not necessarily easy. You can identify where tumors are, and this information, feature extraction from images, is very powerful. It is also the only data modality that is used longitudinally today in cancer to monitor response to treatment.

Another modality that will become important is liquid biopsy testing to follow patients longitudinally, based on molecular profiles and minimal residual disease (MRD). If you think about computational pathology, H&E staining in particular will be important. I am more skeptical about immunohistochemistry at scale, given feedback from pathologists; multiplexing may introduce too much signal and create confusion. Proteomics has potential, but clinically, it is not quite there yet. Even the most advanced actors are not fully at clinical utility.

Over time, we will need to combine these modalities and apply smart algorithms to extract signals and support decision-making. In the end, this is what matters: not computing data unless it brings value to the oncologist, pathologist, biologist, or geneticist.

 

IPM: How is the SOPHiA interface designed for clinicians in practice? What does the user experience look like across different use cases, such as oncology or liquid biopsy workflows?

Camblong: It is a web-based interface you log into. For example, if you are at Moffitt Cancer Center in Florida, using the platform for hematological malignancies, you will see which mutations are detected with high sensitivity and how actionable they are. If you are in a hospital in the U.K. using it for liquid biopsy testing, you will see the mutations identified for those patients.

We also have customers using it from a multimodal perspective, more from an oncologist’s point of view, where they can see how similar patients with similar molecular profiles respond to treatments elsewhere. For us, this includes partnerships with major clinical genomic databases. Through these, we provide access to additional data layers for institutions, even when the patient data originates locally.

The interface is always web-based. In the backend, we use microservices to compute data using AI, deep learning, machine learning, statistical inference, and pattern recognition. The user then leverages this information to make decisions and answer clinical questions.

 

IPM: Given the diversity of data sources and technologies, how do you approach standardization and harmonization across datasets, particularly in a global context?

Camblong: We operate in over 70 countries. We support local data production and management, but within a framework of collective knowledge. It is important to align solutions with regulations. In some countries, we operate in research mode only. In Europe, some applications are IVD, and in the future possibly In Vitro Diagnostic Regulation (IVDR) or companion diagnostic solutions.

The key is to build technology with optionality, documenting how it is built and its intended use. If you want to make clinical claims, you must conduct clinical studies. The foundation is design control, like in aviation, so that you ensure sensitivity, specificity, reproducibility, repeatability, and robustness, regardless of regulatory frameworks.

 

IPM: How does your platform adapt to the wide variety of user systems, including different sequencing instruments, workflows, and laboratory environments?

Camblong: The backend is fully engineered and automated. But workflows differ across hospitals due to global constraints and complexities. Managing this heterogeneity while delivering consistent outputs means adapting to different workflows. This is not easy, but we have demonstrated strong performance. For example, with Memorial Sloan Kettering, we accessed both their data and their applications, MSK-IMPACT and MSK-ACCESS. We industrialized these within SOPHiA without infringing on IP, enabling hospitals to produce data locally and leverage our algorithms. We achieved over 98% concordance across sites, comparable to repeating sequencing within a single workflow.

We also work with multiple sequencing vendors to ensure compatibility across instruments and consumables. Because we process large volumes of data, we can also advise on optimal workflows for specific applications. Since we are paid per use, our incentives are aligned with hospitals; better workflows mean more patient cases and better outcomes.

On AI: it is a toolbox. Different models suit different problems. Large language models are useful for text and sometimes images, but not everything. Understanding biology and data diversity is key to selecting the right mathematical model that scales effectively.

 

IPM: As you expand into adjacent domains like radiology, how do you approach entering new clinical areas while ensuring relevance and usability?

Camblong: Always with partners, healthcare institutions. We are strong in software, AI, and biology, but not medical practice. We co-develop with clinicians to ensure integration into workflows and real clinical benefit. For example, with MD Anderson, we collaborate on translational and routine lab work to move technologies into clinical practice, such as transcriptomics for cancer subtyping and MRD.

In multimodality, we work case by case. For instance, in kidney cancer in France, we partnered with the UroCCR network, analyzing 27,000 patient cases. This allowed us to identify signals and predict responses to immunotherapy. Innovation only matters if it is adopted in practice.

 

IPM: How actionable are your clinical decision-support tools today, and how do you incorporate real-time or longitudinal data?

Camblong: It depends on regulations. In some places, like the U.K., the platform provides information to oncologists, who then interpret it. For multimodality, feedback loops are essential, linking molecular data, treatment, and outcomes.

With UroCCR, we continuously improve algorithms using real-world data. We should be leveraging post-market data more systematically to refine treatment decisions. Real-world complexity can reveal which patients truly benefit from therapies. Longitudinal data is critical, not just for outcomes, but also for avoiding adverse effects. For example, some ovarian cancer patients benefit from PARP inhibitors but may develop leukemia. Understanding these patterns requires real-world data loops.

 

IPM: How do you think about data ownership, access, and control?

Camblong: Ownership does not exist in a strict sense. Individuals are the ultimate controllers. Hospitals and companies are processors. Data is critical for AI, but our model is decentralized: hospitals retain control of their data. Algorithms learn from data, but once trained, they can deliver insights without retaining raw data, enhancing privacy.

Also, oncology data does not age well because treatments and technologies evolve rapidly. What matters is continuous exposure to new data. Collective intelligence through networks and platforms is essential for precision medicine.

 

IPM: How does SOPHiA approach cross-border collaboration and democratization?

Camblong: Democratization means making technology accessible and usable. For example, in India, a hospital previously sent samples to the U.S., with high costs and six-week turnaround times. We enabled local testing within months, reducing turnaround to under two weeks and building internal expertise. This increased testing volumes and improved clinical adoption.

 

IPM: Are there areas less amenable to your approach?

Camblong: About 80% of our work is in cancer, 20% in rare disorders. Rare diseases require even more collaboration due to limited data. We support peer networks where clinicians share insights, for example, variant classifications, helping others make faster decisions. As medicine becomes more precise, collaboration becomes even more critical.

The post Jurgi Camblong: Data-Driven Doctors Without Borders appeared first on Inside Precision Medicine.