The Opportunity
Cancer is a complex and multifaceted disease and remains one of the most formidable challenges in modern medicine. Despite remarkable advancements in treatment over the past few decades, traditional approaches, such as chemotherapy, radiation, and surgery, typically operate on a one-size-fits-all principle. While these approaches can be effective in many cases, they can also lead to significant side effects and varying degrees of success among patients. This variability underscores the urgent need for more precise, individualised treatment strategies, a need that has given rise to the field of personalised cancer treatment.
Antibody-Drug Conjugates (ADCs) in oncology represent a significant advancement in cancer treatment, offering a promising new avenue for targeted therapy. The potential of ADCs in oncology is immense, as they can be tailored to target a wide range of tumour antigens, making them versatile tools in treating various cancers.
The Challenge
The response to Antibody-Drug Conjugates (ADCs) in oncology varies among patients, highlighting the complexity of cancer biology and the challenges of personalised treatment. While ADCs selectively target and destroy cancer cells there are other factors that influence their effectiveness. There are other challenges with ADCs, for example production is a complex task with high costs, the impact of which is broad. The result is only around 10% of potential ADCs have made it through clinical trials. In other words, ADCs need greater success in clinical studies to become a truly transformative treatment for cancer patients globally.
The Gap
Despite advances in research, we do not understand why ADC treatments fail. Filling this gap in our knowledge is essential to helping more ADCs reach their clinical trial endpoint.
It has been known for decades that bacteria are present within tumours. The BioCorteX Carbon Mirror platform suggests that bacteria that reside inside the tumour microenvironment have the potential to interact with ADCs. This discovery fills a significant gap in pharmaceutical companies’ understanding of why clinical trials fail and, equally importantly, why approved treatments do not benefit everyone.
Revolutionary Science
At BioCorteX, we have advanced a scientific approach labelled ‘Unified Biology’. We understand that creating novel technology is valuable but will only take you so far. It is the combination of our innovative technology with a ‘unified biology’ scientific approach that makes BioCorteX unique.
Conventional approaches to biology are reductionist, narrow and siloed. They typically test a single pathway and assume a linear response.
At BioCortex, we pioneer a broader approach that unifies biology by considering the full range of high dimensions in biological data. We believe that understanding the shape and non-linear interaction within data is essential in bridging the gap between the bench and the bedside. We can move seamlessly between representations of cells to an individual’s data and finally to large populations with key emergent properties and insights appearing at each level.
By leveraging a unified biology approach, BioCorteX makes it easier to identify gaps in disease understanding, connecting insights across multiple biology dimensions, thus preventing siloed one-dimensional approaches.
Supercharged with Technology
To meet the complex scientific challenges and implement a unified biology approach, BioCorteX technology needs to work at extraordinary scale and speed enabling scientists and clinicians to design hypotheses and deliver results in hours or days and not weeks or months.
The scalability of the BioCorteX technology platform across higher dimensional data and emergent behaviour allows our scientists to separate signal from noise. Not having to compromise on scale coupled with being able to work incredibly fast supports iterative hypothesis testing and rapid development. This is a key to success when teams are exploring uncharted areas of science and technology.
Carbon KnowledgeTM achieves these requirements by publishing a new version every single day. This is a huge undertaking when there are over 16 billion edges, or relations between nodes in the graph, and growing. While this scale and speed is impressive, efficiency is equally vital for a startup. The cost of these releases typically is less than $30 in Carbon Knowledge. This ability to combine huge scale, truly enabling speed, and low costs is helping our team explore more novel hypotheses and achieve results faster than they could anywhere else.
Trust is just as important as scale and speed. Due to the way Carbon Knowledge is architected within Google Cloud, full transparency is achieved, with the user being able to track data back to the source. When coupled with a robust science-ready data health check, confidence in an analysis is quickly established. This architecture enables a BioCorteX scientist to verify data reliability in minutes and make an informed decision immediately on the feasibility of a potential study.
ADC Case Study Example Tech Stats for Search Performance
| Metric | Value |
|---|---|
| Knowledge Graph Search Space | > 435 Terabytes |
| Number of edges | > 16 Billion |
| Typical curated snapshot | > 1.2 Terabytes |
| Typical ADC query size | ~ 5 Gigabytes |
| Average query execution time | < 10 Seconds |
Carbon Knowledge provides BioCorteX scientists with the ability to seek answers to extremely complex questions. With a search space of almost half a petabyte and growing, they are able to design and run highly contextually relevant hypotheses in seconds with data that they can trust and interrogate back to the source to ensure complete understanding.
// From the original blog:
How BioCorteX uses BigQuery to help answer the question: “Will this drug work?”
With the help of Google Cloud, BioCorteX is changing the biotech landscape, achieving fast results that would have seemed impossible only a few years ago. BioCorteX has developed two complementary technologies that are accelerating this change. The first is called Carbon Knowledge™, an extraordinarily large biology-based knowledge graph with over 3bn nodes and 16bn edges, stored on BigQuery. The second is the foundational emulator, Carbon Mirror™, which builds on Carbon Knowledge™ to uncover hidden interactions between bacteria, the human physiology, diseases, and drugs. Actionable insights from Carbon Mirror are already informing deep mechanistic understanding and greater certainty of treatment response to enable the accelerated development of sophisticated drugs at scale.
