We harness breakthrough generative AI to design entirely novel proteins and antibodies from scratch, targeting historically undruggable biological mechanisms. Our foundation models create therapeutic molecules that nature has never seen—opening new frontiers in precision medicine and transforming what's possible in drug development.
For decades, approximately 85% of the human proteome has remained "undruggable"—proteins and biological targets that conventional drug discovery methods simply cannot address. These include intrinsically disordered proteins, protein-protein interactions, and transcription factors that drive some of humanity's most devastating diseases, from aggressive cancers to neurodegenerative conditions.
JACKSTAR was founded to fundamentally change this paradigm. By combining cutting-edge generative artificial intelligence with massive-scale proteomics data generation, we've created a platform capable of designing therapeutic molecules that evolution never produced—proteins and antibodies engineered from first principles to bind targets that have eluded pharmaceutical research for generations.
"We're not optimizing existing molecules. We're creating entirely new classes of therapeutics that expand the boundaries of what medicine can achieve."
Our foundation models, trained on billions of protein sequences and structural data points, can predict molecular behavior with unprecedented accuracy. More importantly, they can generate novel molecular architectures designed for specific therapeutic functions—creating medicines that are purpose-built rather than discovered by chance.
Backed by one of the largest initial funding commitments in venture history and guided by a leadership team of former pharmaceutical executives and Nobel laureate advisors, JACKSTAR represents a new era in drug development. We're building the future of precision medicine, one designed molecule at a time.
Our end-to-end platform combines proprietary foundation models, high-throughput experimental validation, and advanced computational biology to accelerate the journey from target identification to clinical candidate. Every component is designed to work in concert, creating a system more powerful than the sum of its parts.
From novel target identification to optimized lead candidates, our integrated capabilities address every stage of early drug discovery with unprecedented speed and precision.
Identify previously undruggable targets through multi-omic data integration and AI-powered analysis of disease mechanisms. Our models reveal hidden vulnerabilities in complex biological systems that traditional approaches miss.
Generate entirely novel therapeutic molecules—proteins, peptides, and antibodies—designed from first principles to engage specific targets with optimal binding and selectivity characteristics.
Predict protein structures with atomic accuracy, including challenging targets like membrane proteins, multi-domain complexes, and the conformational ensembles critical for drug binding.
Systematically optimize binding affinity through AI-guided sequence evolution. Achieve picomolar binding while maintaining selectivity, stability, and manufacturability.
Predict and optimize manufacturability properties from sequence alone. Screen for aggregation, immunogenicity, stability, and expression yield before synthesizing a single molecule.
Forecast in vivo pharmacokinetics, toxicity risks, and efficacy signals using our integrated biological simulation models trained on comprehensive preclinical datasets.
Our internal pipeline focuses on targets that have resisted conventional drug discovery approaches—proteins and pathways where traditional small molecules and standard biologics have consistently failed. By designing molecules from first principles, we're unlocking therapeutic opportunities that represent significant unmet medical need across multiple disease areas.
Across eight therapeutic areas, our AI platform is generating novel candidates against 47 distinct targets, with multiple programs advancing toward clinical development. Each program leverages our full technology stack: generative design, structure prediction, affinity optimization, and developability assessment working in concert.
Designing degraders and binders for transcription factors and oncogenic protein-protein interactions that drive tumor growth and therapeutic resistance.
Engineering precision modulators of immune checkpoints and inflammatory pathways for autoimmune and inflammatory conditions.
Developing brain-penetrant biologics and aggregation inhibitors for neurodegenerative diseases with no current effective treatments.
Creating broadly neutralizing antibodies and pan-viral therapeutics designed to counter emerging pathogens and antimicrobial resistance.
Engineering next-generation metabolic modulators targeting obesity, diabetes, and NASH with improved efficacy and safety profiles.
Developing therapeutics for heart failure, atherosclerosis, and hypertension targeting previously intractable pathways.
Creating protein replacement therapies and corrective biologics for rare diseases caused by protein deficiency or dysfunction.
Engineering long-acting biologics for retinal diseases with extended durability to reduce treatment burden for patients.
Our integrated platform compresses timelines that traditionally took years into months, accelerating the path from biological insight to clinical candidate through continuous AI-driven optimization.
Deep computational analysis of target biology, structural characteristics, and disease relevance to define the optimal therapeutic hypothesis and druggability assessment.
AI foundation models generate thousands of novel molecular candidates optimized for target engagement, selectivity, and therapeutic properties from the first design iteration.
High-throughput experimental platforms synthesize and characterize top candidates, feeding results back into model optimization for continuous improvement cycles.
Iterative refinement of leads for affinity, selectivity, stability, and developability to deliver clinical-ready candidates with optimized drug-like properties.
Our core technology builds upon the transformer architecture that revolutionized natural language processing, adapted and extended for the unique challenges of biological sequence understanding. Just as large language models learn the grammar and semantics of human language, our models learn the fundamental principles governing how amino acid sequences fold into functional proteins. Trained on billions of evolutionary sequences and hundreds of millions of structural annotations, these models capture the deep patterns of protein biology—enabling both accurate prediction and creative generation of novel molecular designs that have never existed in nature.
Moving beyond the traditional paradigm of screening existing molecular libraries, our generative approach designs molecules from first principles. Given a target binding site or desired therapeutic function, our models generate sequences optimized to achieve that goal—creating proteins and antibodies that have never existed in nature. This approach unlocks therapeutic modalities that evolutionary selection never explored, opening new possibilities for treating diseases that have resisted conventional drug discovery for decades.
Understanding how designed molecules will behave in complex biological systems requires modeling at multiple scales—from atomic interactions to cellular pathways to physiological outcomes. Our platform integrates physics-based molecular dynamics with learned models of biological behavior, enabling prediction of binding kinetics, cellular uptake, tissue distribution, and therapeutic efficacy. This multi-scale approach allows us to optimize for real-world performance rather than isolated biochemical metrics that often fail to translate to clinical success.
The true power of our platform lies in the tight integration of computation and experimentation. Every molecule synthesized and tested generates data that improves our models. Our automated wet lab platforms can characterize thousands of variants per week, creating a continuous feedback loop that rapidly refines predictions and enables active learning strategies that focus experimental resources on the most informative experiments. This iterative approach compounds improvements exponentially, with each cycle making the platform more accurate and more efficient.
Our proprietary proteomics capabilities generate training data that doesn't exist anywhere else. Using advanced mass spectrometry and next-generation sequencing, we characterize protein expression, post-translational modifications, and interaction networks across disease-relevant tissues and cell types. This unique data advantage allows our models to understand biology with a depth and specificity that publicly available datasets cannot provide—giving us a fundamental edge in predicting therapeutic outcomes and designing molecules that work in the complex environment of human disease.
The integration of generative AI with biological understanding represents the most significant advance in drug discovery since the advent of molecular biology. We're not just finding new medicines—we're learning to speak the language of life and write new chapters that evolution never imagined.
Leadership Perspective — On the future of AI-driven therapeutics
Whether you're exploring collaboration opportunities, interested in our platform capabilities, or seeking to address challenging therapeutic targets, we'd like to hear from you.
JACKSTAR is actively pursuing partnerships with pharmaceutical companies, biotechnology organizations, and research institutions aligned with our mission to transform drug discovery through AI. We're particularly interested in collaborations targeting historically undruggable proteins and novel biological mechanisms where our platform can create breakthrough therapeutic candidates.