Predictive Artificial Intelligence and Robotics Platform
The LOCUS platform industrialized the development of precision, engineered bacteriophage therapies. As a precision therapy, our products are built to be broadly applicable across patients, accessible as an off-the-shelf product, and precise enough to reach into the human body and selectively remove a bacterial species without disrupting the microbiome.
The LOCUS platform was created by combining Locus Biosciences’ expertise in bacteriophage synthetic biology with EpiBiome’s pioneering the use of machine learning in combination with high-throughput automation and advanced bioinformatics. With continued investment, we have built one of the most advanced precision drug development platforms deploying AI, automation, and synthetic biology.
Central to our mission is LOCUS, a powerful technology-led platform that enables us to develop products that selectively remove bacterial species at equivalent or superior levels than antibiotics without the drawback associated with broad-spectrum antibiotics. Instead of employing the age-old approach to building phage therapies, LOCUS industrializes the process by deploying a suite of robotic tools, predictive modeling, and world-class manufacturing to build its precision medicines.
The scale of proprietary data that LOCUS generates enables the development of quality machine learning models. In short, better data results in better predictions. With each product developed, LOCUS becomes better at designing the next one. With five active programs advancing into and through the clinic, LOCUS has become one of the world’s most robust bacteriophage development platforms.
The computational powerhouse of LOCUS, the Cloud Core not only stores all the data we generate but enables access and learning from natural language processing (NLP)-inspired models and other predictive models used to rapidly advance our engineered phage cocktails.
To enable global sourcing of bacterial strains, mass discovery of bacteriophages, and rapid engineering of target phages, we utilize automated biobanks to store and track the tens of thousands of bacteria, phage, and environmental samples we have on site.
Home to the suite of sequencing capabilities within LOCUS, the data generated by this core feeds the NLP-inspired models that enable our team to build clinically-relevant bacterial training panels that are genomically diverse and to accurately predict optimal engineering sites within phage genomes.
To enable conducting experiments at a scale that human hands could not accomplish alone, we built high-throughput robotics capabilities that enable our team to test millions of bacteria-phage interactions for each program and to ensure the efficacy seen in vitro translates to the clinic.
We built an award-winning 10,000 sq ft clinical- and commercial-scale cGMP manufacturing facility that meets the specifications of the US FDA, EMA, and Japanese Ministry of Health, providing us with a right-sized facility that produces high potency and purity drug product.
Designing an effective phage therapy starts with optimizing the clinical panel. This ensures that the bacterial strains tested are representative of those that will be encountered in the clinic. This is why we build clinical panels of hundreds of bacterial isolates that are bioinformatically representative of up to 10,000 publicly available genomes.
Building a fixed phage cocktail that rivals the efficacy of antibiotics requires phage discovery at a scale that has never been attempted before. We discover thousands of phages for each program to feed our cocktail optimization engine.
Automated banking provides an unprecedented level of control, supporting rapid and accurate creation of bacterial strain panels. Barcoded and long-read sequence-validated bacterial isolates ensure that panels are repeatably assembled to consistent specifications. When testing millions of phage-bacteria interactions, having automated processes that maintain the integrity of the panels is essential.
Robust, comprehensive characterization of both isolated phages and target bacteria generates critical genotypic and phenotypic data. These data feed predictive models that drive phage selection, cocktail optimization, and phage engineering.
To test every isolated and characterized phage in a cocktail of six phages would require trillions of permutations. Even automated robotics can’t deliver that scale, so we turn to in silico computational models that predict the highest-performing cocktails. By leveraging these models, we to turn an impossible challenge into a manageable task.
Computational models are excellent at predicting the highest-performing cocktails, but they still must be evaluated. Testing these cocktails across a variety of conditions ensures that we select cocktails in which all of the phages complement, rather than antagonize, each other, allowing robust bactericidal activity and mitigation of resistance development.
Millions of years of co-evolution has produced an ecological balance between phage and their bacterial prey. Synthetic biology allows us to tip the balance in favor of phage killing the target bacteria by enhancing phage with bactericidal payloads.
We built an award-winning manufacturing facility that allows us to strategically control the manufacturing of our products, enabling us to move into and through clinical development at a lower cost and a faster pace than we could if we were relying on external manufacturing partners.
If you are interested in working with us on developing precision medicines, please reach out below.