Overview
My chat with Ron Alfa, MD, PhD, veered into the essence of diseases. Some cancers are defined by their genes, making it easier to create precise treatments and sort patients. But in immuno-oncology, those clear lines haven’t been drawn yet. Alfa pointed out a significant gap: we can’t clearly define patient groups by their immune biology. As a result, promising new immuno-oncology molecules haven’t been as plentiful as one might hope.
The reason? It comes down to not really knowing which tumors will respond to different treatments. Alfa said we lack a system to tell us that some tumors suppress the immune system using mechanism A, others mechanism B, and so forth. If we could sort that out, we could kick immuno-oncology into high gear and start classifying tumors by their immune resistance mechanisms.
Genomics doesn’t map neatly to immune biology. So, Alfa and Jacob Rinaldi, PhD, decided to tackle this problem by founding Noetik. They met at Stanford and now lead a company focused on understanding the right data to figure out tumor biology and leverage machine learning.
No ordinary spatial company
Noetik might not be a spatial biology company, but it uses spatial techniques heavily. Alfa mentioned they’re the biggest users of two new spatial platforms (one for proteomics, another for transcriptomics). For Noetik, spatial data is the key for machine learning models. They handle vast amounts of data differently than academic researchers. Their business model needs them to scale up, running machines 24/7, uploading data to the cloud, and maintaining quality with industrial-level standards.
In just six months, they gathered data from hundreds of patients, including genomic data, H&E staining, proteomics, and transcriptomics. The machine learning models train on this wealth of data. This step could lead to understanding tumor biology in ways humans can’t achieve—it’s about classifying tumors into subsets based on their immune properties. They employ self-supervised learning, allowing models to independently identify data patterns. The aim? Discover novel biology that hints at what kind of cancer drugs to develop.
Industry challenges and frustrations
Alfa’s frustration with existing cancer drugs drove much of his work. Despite significant biological knowledge, effective cancer drugs remain elusive. This irritation fueled his career, leading him and Rinaldi to co-found Noetik. Just over a year old, the company already has footprints in South San Francisco and Boulder, thanks to an oversubscribed $14M seed round.
The fundamental problem Noetik wants to solve is delivering better cancer treatments. Their unique angle involves using different data to crack the code. The company’s name, Noetik, stems from a Greek word meaning intellectual. It reflects their ambition for artificial intelligence to decode the intricacies of patient groups, potentially paving the way for improved cancer therapeutics.
A new approach to drug discovery
Noetik is eyeing a future where their models help create a portfolio of drugs tailored for specific tumor types. They start with patients, employing reverse translation to guide drug development. Their platform doesn’t predict molecules or design drugs directly but aims to reveal new biology, informing what drugs to make eventually.
Here’s a brief snapshot of Noetik’s focus:
- Immuno-oncology: Tackling tumor and immune system interactions
- Spatial biology: Using spatial data to train machine learning models
- Machine learning: Developing self-supervised models
- Cancer therapeutics: Aiming for more effective cancer drugs
Noetik’s innovative approach could become a blueprint for other companies dealing with drug discovery and cancer treatments. As they continue to explore the intersection of spatial biology, machine learning, and tumor biology, the future of precision oncology looks promising.
Precision immunotherapy and beyond
Eventually, Noetik could transform immuno-oncology into precision oncology. By understanding how different tumors interact with the immune system, they could classify tumors in new, therapeutically relevant ways. Artificial intelligence plays a crucial role, with models deciphering complex datasets to uncover insights.
The company’s long-term vision includes developing precision immunotherapies. Noetik’s machine learning models might one day identify which tumors are most likely to respond to specific treatments, revolutionizing how we approach oncology therapies.
When I reflect on Noetik’s journey, it’s clear their blend of innovation and data-driven methods sets them apart. Combining machine learning, spatial biology, and a strong focus on immune biology, they might just be on the cusp of something groundbreaking in the fight against cancer.
Leave a Reply