Thanks for joining this newsletter. If I've added you without asking you, it's because I think you're smart and I value your opinion and feedback. Feel free to unsubscribe.
In this weekly roundup, I explore all things at the intersection of cancer research and artificial intelligence. It's my personal exploration as I figure out how I can help beat cancer.
This week I want to focus on Atomic AI, a biotech startup that announced a Series A funding a few days ago.
Atomic AI raises $35 million to help discover new drugs with AI.
Atomic AI was founded by Raphael Townshend in 2021, the year he finished his computer science PhD at Stanford. The company previously raised a $7M seed and was operating in stealth mode until now.
A simplification would be to say that Atomic AI is doing for RNA what Deepmind's Alphafold did for proteins, in an effort to find novel drugs.
What's RNA? RNA carries out the instructions encoded in our DNA. The RNA copies our DNA to make proteins. In short, DNA makes RNA makes proteins. Also, many viruses don't contain DNA and their genetic information is stored in the RNA. These are called "RNA viruses" and include measles, rubella, ebola, small pox and coronaviruses like Covid-19.
A lot of academic research has been done on DNA and proteins. The pharma industry has focused on proteins as the mechanism for drugs. Many drug discovery programs are dedicated to finding molecules capable of binding reliably and exclusively with disease-causing proteins.
But this focus on proteins has led to some dead ends. Some proteins are, so far, impossible to drug, resulting in illnesses resistant to medication.
The core idea behind Atomic AI is that RNA offers a whole new target to go after. Instead of going after proteins, Atomic AI goes upstream to the RNA that creates the protein. This is captured in the title of Techcrunch's article: "Atomic AI envisions RNA as the next frontier in drug discovery".
There is a lack of knowledge around RNA structures. This is why Atomic AI has developed a machine-learning model to predict the 3D structure of RNA, find how to target it and what molecule could bind with it.
Atomic AI is significantly expanding the surface area that can be targeted to treat diseases.
In terms of business model, Atomic AI could have licensed out its core AI product as a "RNA structure as a service" platform for others to use. Instead it is investing it's own wet lab and drug discovery program.
As reported by MedCityNews, "Data generated by the wet lab are used to train the AI to discover novel targets on the three-dimensional structure of RNA, Townsend said. The AI makes predictions that inform additional wet lab experiments. Those results feed additional AI analysis, continuing a virtuous cycle."
The company won't disclose which diseases it is going after yet but has noted that it would seek to develop molecules for use in cancer, neurodegenerative disorders, cardiology, rare disease, and infectious disease.
The founder, Raphael Townshend, is only 30 years old. Digging into his CV, I noticed he did a summer internship at Google's Deepmind in London in 2019, where he worked on Alphafold.
The world needs more visionary risk-takers like Raphael Townshends.
I'm rooting for you, Raphael.
Who am I?
I'm currently fighting blood cancer (Non-Hodgins lymphoma). I've decided to dedicate my life to advancing cancer treatment. This newsletter is my personal research. You can find out more about my mission here. Get in touch if you want to connect and potentially collaborate.