Y Combinator · Video summary

This Startup Wants To Catch Cancer Before It Spreads

Billion to One uses a novel DNA amplification noise reduction technique to detect rare genetic markers, initially for prenatal testing and now targeting early-stage cancer detection.

Y Combinator · April 6, 2026 · 21-min episode · 5-min read

This Startup Wants To Catch Cancer Before It Spreads

Key takeaways

💡Detecting rare DNA fragments (e.g., one base pair out of billions, or trace amounts of fetal/tumor DNA) requires overcoming significant amplification noise; Billion to One adds synthetic DNA to quantify and subtract this noise.00:12
🚀Consider how noise reduction techniques can be applied to other data-intensive problems where signals are faint or obscured.03:51
🎯The core innovation is transforming a complex biological problem (detecting dilute DNA) into a computational problem by accurately measuring and correcting for errors introduced during DNA amplification.04:26
🔍Early commercialization challenges, like having only one physician user months after launch, were overcome by direct-to-patient marketing and empowering patients to convince their doctors.08:05
📈Leveraging AI and computer vision significantly accelerated bottleneck processes, such as sample accessioning, reducing a 60-second manual task to near-instantaneous automated handling.10:30
💡Explore opportunities to integrate AI and computer vision to automate and accelerate bottleneck processes in data handling or sample management.10:30

Chapters

0:01🎯 Introduction to Genetic Testing and the Needle in a Haystack Problem
0:35🎯 Current State of Prenatal Testing and Future Goals
1:05🎯 Founders and Company Background
1:44🎯 What Billion to One Does
2:28🎯 Scale and Impact of the Company
3:13🎯 Key Insights Enabling Early Detection
4:46🎯 Technical Approach to Detecting Tiny DNA Fragments
5:17🎯 Rapid Development and Commercialization
9:57🎯 Laboratory and Processing Workflow
12:00🎯 Sample Multiplexing and Computational Analysis
12:56🎯 Expanding from Prenatal to Cancer Detection
18:14🎯 Future Goals and Step-by-Step Strategy
19:02🎯 Challenges and Why No One Else Has Achieved This
20:12🎯 Commitment to Impact and Company Culture

The digest

Billion to One, a pioneering molecular diagnostics company, has transformed prenatal genetic testing and is now aggressively advancing early cancer detection using blood-based DNA analysis. Founded by two PhD students in 2017, the company leverages groundbreaking technology to identify minuscule DNA fragments in the bloodstream, enabling detection of genetic diseases in fetuses and cancerous mutations at stages previously undetectable. This stepwise approach—from prenatal testing to late-stage and then early-stage cancer detection—addresses one of medicine’s hardest challenges: finding a "needle in a haystack" of billions of DNA base pairs with remarkable sensitivity and specificity.

Core Technology and the Needle-in-a-Haystack Challenge

Billion to One’s core innovation lies in detecting tiny amounts of cell-free DNA shed by tissues—including fetal and tumor DNA—into the bloodstream. Each human genome contains about 3 billion base pairs, and diseases like sickle cell anemia or cystic fibrosis typically differ by just a single base pair. Identifying one unique mutation among billions of bases is the "billion to one" problem.

Most existing genetic tests amplify all DNA, which increases not only the signal but also noise—random errors introduced during DNA amplification (PCR). Billion to One’s breakthrough came from introducing synthetic DNA molecules into the sample before amplification. These synthetic templates act as internal standards, enabling the company to track amplification bias and errors at each genomic location. By applying machine learning algorithms, Billion to One subtracts this noise computationally, leaving a clear, accurate signal of the original DNA sample. This converts an extremely difficult biological problem into a manageable mathematical one.