Digital Twins: Drug Discovery’s Next Leap?

Basic Concept of a Digital Twin in the Context of Humans

A digital twin is a virtual model that mirrors a physical entity, such as the human body or its parts, with real-time data exchange

Key Concepts

Definition and Structure: A human digital twin is a virtual model with real-time, bidirectional links to the physical entity, comprising the physical body, its digital representation, and data connections. This structure allows the digital twin to mirror changes in the physical entity and vice versa, enabling dynamic updates.

Types of Human Digital Twins:

• Whole Body Twins: Represent the entire human body, capturing overall health and physiological states
• System or Function Twins: Focus on specific systems, like the digestive system
• Organ Twins: Model individual organs, such as the liver or heart, for organ-specific analysis
• Cellular, Subcellular, or Molecular Twins: Zoom in to finer levels, analyzing cells, organelles, or molecules
• Disease-Specific Twins: Created for specific conditions, like a liver with non-alcoholic fatty liver disease. Composite Twins: Integrate multiple types for comprehensive modeling.
• Reference or Proto-Twins: Serve as templates for building individualized twins
• Levels of Digital Twins: Vary in sophistication based on detail, real-time update frequency (via clinical sensors), and embedded intelligence (data and algorithms) thereby requiring higher fidelity and frequent updates to enhance accuracy
• Digital Twin Instances and Aggregates: Instances: Identical copies for in silico testing, comparing treatment scenarios for individuals. Aggregates: Collections from different individuals for population-level analysis, useful in public health.
• Digital Thread: A temporal data pipeline tracking changes over time, enabling longitudinal analysis and predictive modeling, such as from birth to death for an individual

Potential Applications

• Personalized Medicine: Digital twins enable tailored treatments by modeling individual patient data, such as genetics and lifestyle, to predict treatment outcomes. For example, they can simulate how a patient responds to a drug, reducing trial-and-error
• Precision Public Health: Digital twins track and manage disease outbreaks at population levels, such as during COVID-19, using systems integrating patient data, AI, and blockchain for secure sharing, providing insights for public health planning, simulating intervention effects on communities or cities
• Repository for clinical trials: The human digital twins' bank acts as an organized repository of digital twin instances and aggregates, with version control, potentially for clinical trial matching and interoperability with other banks, potentially speeding up medical research by matching patients based on digital twin characteristics

The Evolution of Digital Twins: From Factory Floors to Biopharma Frontiers

• Digital twins were born in the industrial world - from virtual jet engines or factory assembly lines, where engineers could test designs and predict failures without touching a wrench
• These models, fueled by real-time data and advanced analytics, saved billions and redefined efficiency
• Now, biopharma is borrowing this playbook, adapting digital twins to mirror the complexity of biological systems. In drug discovery, the journey began with computational models which were rudimentary simulations of molecular interactions
• But today, powered by artificial intelligence (AI) and vast datasets like genomics and proteomics, digital twins have evolved into sophisticated virtual representations of cells, organs, and even patients. In therapy area management, they’re shifting us from one-size-fits-all treatments to precision medicine, where every decision is data-driven and patient-specific

Digital Twins in Drug Discovery: De-Risking the Pipeline

• Drug discovery holds the central role in a Biopharma's discovery and yet there has been no way to master it It comes with high stakes, higher costs, and ~90% failure rate in clinical trials
• Digital twins are here to change that, offering virtual sandboxes to test hypotheses, predict outcomes, and optimize candidates before we spend a dime on wet labs.

Simulating Biology at Scale

Picture this: a digital twin of a human cell, built from single-cell RNA sequencing data, reacting to thousands of drug candidates in silico. No pipettes, no animal models, just pure computational power and it is just the beginning!

Trailblazers in the Digital Twin Space for BioPharma

Somite Therapeutics' AI-driven digital twins of embryonic cells simulate how stem cells differentiate into new cell types

• By identifying novel protocols faster than traditional methods, Somite is accelerating target discovery and cutting months off early-stage development & it can contribute a significant head start on rare disease therapies.

DeepLife takes a broader approach, crafting digital twins of human cells to evaluate drug responses across disease states

• Their platform, rooted in omics data, predicts efficacy and toxicity, speeding up candidate selection and even repurposing existing drugs

Companies Utilizing Digital Twins in Drug Discovery -Case Studies

• A growing number of innovative companies are leveraging digital twins to enhance drug discovery, clinical trials, and therapeutic production. Below, we explore key players, their approaches, and their contributions to the biopharma industry, emphasizing how digital twins deliver commercial value and support drug development

Aitia: Pioneering Causal AI with Gemini Digital Twins

Overview:

• Aitia (formerly GNS Healthcare) is a leader in applying digital twins to drug discovery through its Gemini Digital Twins technology
• These computational models simulate disease biology by integrating genetic, molecular, and clinical data to reveal causal mechanisms driving health outcomes

Approach:

• Aitia’s proprietary AI platform, REFS (Reverse Engineering and Forward Simulation), uses causal AI to construct digital twins that go beyond correlation to identify cause-and-effect relationships

Rather than merely noting a gene’s association with a disease, Aitia’s twins determine whether it drives the condition -an insight critical for target validation

Partnerships

• Aitia partnered with Servier to accelerate drug development - This partnership was form with Servier trying to advance a Parkinson’s disease therapy targeting the leucine-rich repeat kinase 2 (LRRK2) protein, implicated in disease progression
• Implementation: Aitia’s Gemini Digital Twins integrated multiomic data, from thousands of Parkinson’s patients to model disease biology. Its Gemini Digital Twins simulate gene and protein knockdowns in silico, predicting their impact on disease progression and identifying novel drug targets.
• The twins simulated disease mechanisms, identifying genetic and molecular profiles linked to progression and responsiveness to Servier’s LRRK2 inhibitor
• Virtual testing pinpointed biomarkers distinguishing responders from non-responders
• Outcome: This stratification enabled Servier to design targeted clinical trials, focusing on patients likely to benefit, reducing trial size and risk.
• Commercial Benefit: Aitia’s digital twins saved Servier months in trial preparation by bypassing broad screening, potentially reducing costs by millions
• The precision in patient selection enhanced trial efficiency, positioning the LRRK2 inhibitor as a frontrunner in Parkinson’s treatment.
• Contribution to Drug Launch: While still in development, the drug’s path to market has been de-risked. The use of digital twins increased the likelihood of clinical success by ensuring trials target the right patients, potentially leading to a successful launch that captures a significant share of the neurodegenerative market.
• Similarly, collaborations in oncology have identified targets for hard-to-treat cancers, enhancing Aitia’s reputation as a high-value partner

Contribution to Drug Launches:

While many of Aitia’s projects are in preclinical or early clinical stages, its digital twins de-risk drug development by improving target selection and trial design, increasing the likelihood of successful launches. The precision and efficiency they bring could shorten development timelines by years, a significant commercial advantage

DeepLife: Digital Twins of Human Cells for Drug Screening

Overview:

• DeepLife is revolutionizing preclinical drug discovery by creating digital twins of human cells - Its platform uses AI to model cellular responses to drugs, predicting efficacy and mechanisms of action

Approach:

• DeepLife integrates omics data -genomics, transcriptomics, proteomics into its digital twins, enabling virtual experiments that mirror real-world cellular behavior
• The platform’s interpretable AI framework provides transparency into prediction rationales, fostering trust among researchers. DeepLife targets diseases like cancer and infectious conditions, where rapid drug screening is critical

Partnerships

• DeepLife aimed to screen small molecule drugs for cancer, using digital twins of cancer cells to predict efficacy

Applications:

• The company’s digital twins evaluate how diseased cells respond to thousands of compounds, identifying promising candidates for further testing
• Proof-of-concept studies have demonstrated their ability to predict responses to cancer treatments and viral infections, validated subsequently in the lab

Commercial Impact:

• By reducing the scope of wet-lab screening, DeepLife slashes preclinical timelines and costs. For instance, narrowing a pool of 100 compounds to 10 viable candidates can save months of lab work and significant expenses, attracting partnerships with biopharma firms seeking to optimize their pipelines

Contribution to Drug Launches:

• DeepLife’s technology has not yet directly resulted in a marketed drug, but its role in identifying high-potential candidates accelerates the journey from discovery to clinical trials & its success in oncology screening could lead to breakthroughs in precision cancer therapies

Somite Therapeutics: Optimizing Cell Therapy with Embryonic Digital Twins

Overview:

• Somite Therapeutics, based in Boston, leverages digital twins to advance cell therapy development, focusing on generating high-quality human satellite cells from induced pluripotent stem cells (iPSCs)

Approach:

• Somite’s digital twin of the embryo integrates AI with rich datasets, such as single-cell RNA sequencing and gene expression profiles, to simulate stem cell differentiation -This virtual model optimizes protocols to enhance cell yield and purity, addressing a key bottleneck in regenerative medicine

Applications:

• In a notable achievement, Somite increased satellite cell purity from 25% to 85% by iteratively testing differentiation conditions in silico. This breakthrough supports the production of cells for therapies targeting muscle regeneration and other conditions

Commercial Impact:

• Somite’s digital twins reduced development time from years to weeks, cutting costs and accelerating the path to clinical trials. The higher cell purity enhances therapeutic efficacy, making Somite’s products more competitive and attractive to investors. This efficiency has bolstered its valuation and partnership potential.

Contribution to Drug Launches:

• Somite’s optimized cells have shown promise in preclinical muscle regeneration studies, positioning the company for clinical trials. The use of digital twins has streamlined this process, increasing the likelihood of a successful cell therapy launch and demonstrating their value in bioprocess innovation

Unlearn.AI: Transforming Clinical Trials

• Overview: Unlearn.AI uses AI-generated digital twins to optimize clinical trials, forecasting patient outcomes and improving trial design
• Approach: The company creates virtual patient models that simulate responses to treatments, integrating clinical and biomarker data. These twins predict trial outcomes, enabling smaller, faster studies without sacrificing statistical power
• Applications: Unlearn.AI has reduced sample sizes by up to 30% in trials for neurodegenerative diseases, shortening durations and costs.

Partnerships;

• Unlearn.AI optimized a Phase III trial for a neurodegenerative disease therapy using digital twins
• Implementation: Virtual patient models simulated trial outcomes, integrating clinical and biomarker data. The twins predicted responses, enabling a 30% reduction in sample size and a shorter trial duration while maintaining statistical power.
• Outcome: The streamlined design preserved rigor, validated against traditional trial data.
• Commercial Benefit: The reduced sample size and timeline saved an estimated $15 million in trial costs, funds redirectable to other projects. This efficiency accelerates time-to-market, a critical edge in competitive fields.
• Commercial Impact: By cutting recruitment needs and timelines, Unlearn.AI saves biopharma partners millions per trial. For a $50 million Phase III study, a 30% reduction translates to $15 million in savings—funds that can be reinvested in R&D. This efficiency accelerates development, enhancing competitiveness
• Contribution to Drug Launch: While the drug remains in trials, Unlearn.AI’s digital twins increase success odds by optimizing patient selection and trial parameters. This de-risking enhances the prospects of a successful launch, potentially transforming neurodegenerative treatment.Its technology ensures rigorous results, validated against traditional trial data. Unlearn.AI’s digital twins de-risk trials, increasing success rates. While specific drugs tied to its technology are still in development, its impact on trial efficiency positions clients for faster, more cost-effective launches, as seen in its neurodegenerative collaborations

Other Notable Players

• Certara: Uses digital twins for pharmacokinetics and pharmacodynamics modeling, aiding dose optimization and reducing clinical trial failures
• Takeda: Explores digital twins in precision medicine, simulating patient responses to tailor therapies, enhancing outcomes in oncology and rare diseases

These companies demonstrate digital twins’ versatility - from target discovery (Aitia, DeepLife) and bioprocess optimization (Somite) to manufacturing (Roche, Biogen) and clinical trials (Unlearn.AI) Their commercial benefits - cost savings, accelerated timelines, and improved success rates, underscore their growing influence in biopharma

Challenges to Mainstream Adoption

Despite their successes, digital twins have not yet permeated the biopharma mainstream due to several hurdles:

• Biological Complexity: Modeling dynamic, heterogeneous systems like cancer or Alzheimer’s requires vast data and computational power, often beyond current capabilities
• Data Integration: Heterogeneous datasets (genomics, clinical records) lack standardization, complicating twin construction
• Scalability: Scaling twins globally demands infrastructure and resources unavailable in many regions
• Regulatory Uncertainty: Evolving frameworks for AI-driven tools create adoption risks
• Cost: High development expenses deter smaller firms, limiting widespread use

The Road Ahead: Challenges and Opportunities

Digital twins are transformative, but they’re not plug-and-play - Here’s what you need to watch:

• Data and Security

These models thrive on data—genomic, clinical, operational. But with great data comes great responsibility. GDPR, HIPAA, and cybersecurity threats loom large.

Comapnies need to have ironclad systems to protect patient privacy and intellectual property

• Regulatory Navigation

The FDA and EMA are warming to digital twins, think the FDA’s exploration of in silico trials, but guidelines are still forming

Early collaboration with regulators will be key to validating your models for approval

Conclusion: Your Digital Destiny

• Digital twins are more than a tool, they’re a paradigm shift. They are changing the way drug discovery is being done, aiming at increased clinical trial success and moving towards precision
• This evolution mirrors biopharma’s own transformation: from trial-and-error chemistry to a data-centric science
• Digital twins are the next leap, promising to cut the $2.6 billion average cost of bringing a drug to market and shrink the 10-15 years it typically takes. For a CEO, that’s not just innovation—it’s a competitive lifeline

Disclaimer: This article represents a fact-based research piece and is not in any manner a guidance document or represents an exhaustive review and is open to inputs from industry experts and a dialogue that will spark meaningful conversations. Examples quoted in this article are not exhaustive but representative of the context.

Few excerpts have been derived from paid financial analysts' report and may not be available for free distribution. In case of any missed or incorrect hyperlinks or incorrect data, please message!

The author represents a CI firm called Intelligience. Please DM her for a detailed discussion around the evolving CI landscape in Biopharma!

Copyright Intelligience 2025 -All rights reserved