Image Credits: Web Summit Biotech & Health How AI is helping solve the labor issue in treating rare diseases

Modern biotechnology can already edit genes and design new drugs, yet thousands of rare diseases still lack effective treatments. According to executives from Insilico Medicine and GenEditBio, the industry’s long-standing obstacle hasn’t been tools or ideas, but a shortage of skilled human labour. Artificial intelligence, they say, is increasingly acting as a force multiplier, enabling scientists to tackle problems that have remained neglected for decades.

Speaking this week at Web Summit Qatar, Insilico president Alex Aliper outlined his company’s ambition to build what he calls “pharmaceutical superintelligence.” Insilico recently introduced its “MMAI Gym,” a system designed to train generalist large language models — including ChatGPT and Gemini — to perform at a level comparable to highly specialised scientific models.

The objective, Aliper said, is to create a multimodal, multitask AI system capable of solving a wide range of drug discovery challenges simultaneously, with accuracy that exceeds human performance.

“We really need this technology to increase the productivity of our pharmaceutical industry and tackle the shortage of labour and talent in that space, because there are still thousands of diseases without a cure, without any treatment options, and there are thousands of rare disorders which are neglected,” Aliper said in an interview. “So we need more intelligent systems to tackle that problem.”

Insilico’s platform processes biological, chemical, and clinical datasets to generate hypotheses about disease targets and potential drug candidates. By automating steps that previously required large teams of chemists and biologists, the company says it can explore enormous design spaces, identify high-quality therapeutic candidates, and even uncover opportunities to repurpose existing drugs — all while significantly reducing development time and cost. For example, Insilico recently applied its AI models toassesse whether approved drugs could be repurposed to treat ALS, a rare and devastating neurological disease.

However, the labour shortage extends beyond drug discovery. Even when AI identifies promising therapeutic targets, many diseases require intervention at a deeper biological level.

GenEditBio represents what some researchers call the second wave of CRISPR gene editing, shifting the field away from editing cells outside the body (ex vivo) and toward precise gene delivery directly inside the body (in vivo). The company’s long-term vision is to make gene editing a single-injection therapy delivered directly to affected tissue.

“We have developed a proprietary ePDV, or engineered protein delivery vehicle, and it’s a virus-like particle,” said GenEditBio co-founder and CEO Tian Zhu. “We learn from nature and use AI machine learning methods to mine natural resources and find which kinds of viruses have an affinity to certain types of tissues.”

The “natural resources” Zhu refers to are GenEditBio’s extensive library of thousands of distinct nonviral, nonlipid polymer nanoparticles — delivery systems designed to transport gene-editing tools safely and precisely into specific cell types.

According to the company, its NanoGalaxy platform uses AI to analyse large datasets and identify how chemical structures correspond to tissue targets such as the eye, liver, and nervous system. The system then predicts how modifying a delivery vehicle’s chemistry can help it deliver genetic material efficiently while avoiding immune reactions.

GenEditBio validates these engineered delivery vehicles through in vivo testing in wet labs, feeding the experimental results back into its AI models to improve predictive performance in subsequent iterations.

Reliable, tissue-specific delivery is essential for successful in vivo gene editing, Zhu said. She argues that GenEditBio’s approach lowers manufacturing costs and standardises a process that has historically been difficult to scale.

“It’s like getting an off-the-shelf drug [that works] for multiple patients, which makes the drugs more affordable and accessible to patients globally,” Zhu said.

The company recently received approval from the U.S. Food and Drug Administration to begin clinical trials of a CRISPR-based therapy for corneal dystrophy.

Combating the persistent data challenge

As with many AI-driven technologies, progress in biotechnology eventually runs into a data bottleneck. Accurately modelling rare and complex aspects of human biology requires far more high-quality data than researchers currently have access to.

“We still need more ground truth data coming from patients,” Aliper said. “The corpus of data is heavily biased towards the Western world, where it is generated. We need to have more efforts locally, to have a more balanced set of original data, or ground truth data, so that our models will also be more capable of dealing with it.”

Aliper noted that Insilico’s automated laboratories generate multilayer biological datasets from disease samples at scale, without human involvement. That data is then fed directly into the company’s AI-powered discovery systems.

Zhu argues that much of the data AI needs already exists within the human genome, shaped by thousands of years of evolution. Only a small portion of DNA directly encodes proteins, while the rest functions more like an instruction set governing gene behaviour — information that has historically been difficult for humans to interpret but is becoming more accessible to AI models, including recent work such as AlphaGenome from Google DeepMind.

GenEditBio follows a similar approach in its laboratory work, testing thousands of nanoparticle delivery vehicles simultaneously rather than individually. The resulting datasets, which Zhu describes as “gold for AI systems,” are used to train internal models and increasingly to support partnerships with external collaborators.

Looking ahead, Aliper says one of the most ambitious goals in the field will be building digital twins of humans to conduct virtual clinical trials — a concept he describes as still being in its earliest stages.

“We’re in a plateau of around 50 drugs approved by the FDA every year, and we need to see growth,” Aliper said. “There is a rise in chronic disorders because we are ageing as a global population … I hope that in 10 to 20 years, we will have more therapeutic options for the personalised treatment of patients.”