Hi, hello!
My name is Sarabeth Schommer, and you just received a brief introduction. I’m going to be talking about automating cfDNA extraction for the RNA-P.

What We’ll Cover Today

I’ll start with an introduction to myself and my company, BillionToOne, and explain what we do. Then we’ll dive into the background and motivation for this project, walk through the development process from an engineering perspective, evaluate performance, and conclude with future improvements.

---

A Little About Me

I know James just gave you a quick intro, but I’ll recap briefly.

I’ve been in the automation industry for a while—almost eight years working with liquid handlers. I got my degree in Biomedical Engineering and started my career at Hamilton as an FAS, where I was exposed to a wide variety of scientific applications and first worked with liquid handling systems.

Later, I joined Mammoth Biosciences, where I applied that experience to product development and became familiar with the Agilent Bravo liquid handler. Around that time, I also began my Master’s in Biotechnology and Management to deepen my understanding of the biotech industry and supplement my hands-on experience.

Now I’m at BillionToOne, which—yes—also has a lot of liquid handlers, including the Lynx.

---

Quick Overview of BillionToOne

We’re a screening company with two divisions: prenatal and oncology. As James mentioned, I’m in the prenatal division, but I’ll touch briefly on oncology.

In oncology, we have two liquid biopsy products:

• Northstar Select

• Northstar Response

In prenatal, our flagship product is Unity Complete. It’s a comprehensive screening test for pregnant individuals, providing both maternal carrier status and fetal risk. As early as nine weeks, we can screen using just a blood tube.

---

Project Focus: cfDNA Extraction

The project I’m discussing today is part of our aneuploidy screening, which determines if there’s an abnormal number of chromosomes.

This is a unique workflow in our CLIA-certified lab.

---

Workflow Overview

We receive blood tubes from clinics and go through accessioning—which ensures the sample is properly labeled and validated. The sample is then separated into components:

• Plasma (3.5–4 mL)

• Buffy coat

For cfDNA extraction, we use the plasma. We start with 24-well format and end in a 96-well format with low-volume elution, which then feeds into PCR and sequencing prep (which we won’t cover today).

While cfDNA extraction is already automated in our lab, it’s currently done on a different platform. With increasing demand and record-breaking test days, we’re looking for a more scalable solution—one that offers:

• Higher throughput

• Better integration

• Customizability

• Lower COGS

• Greater robustness and yield (since cfDNA often has low concentration due to small fragment sizes)

---

Development Process

We’ll break this down into three parts:

1. Configuration

2. Specialty Consumables

3. Method Development

Configuration

We selected the Dynamic Devices LM1800 due to its high-volume capacity, 66 deck positions, and multi-arm capability.

We configured one arm with the MCPA tool, which allows on-the-fly interchangeable heads:

• 24.5 mL head: Great for handling large plasma volumes and bead binding.

• 24.1 mL head: Ideal for transferring samples from 24-well to 96-well format.

• 24-magnet rod head: Useful for low-yield extractions (more on this shortly).

The second arm is the VVP head, Dynamic Devices’ signature pipetting tool, which we use for 96-well plate handling.

For incubation, heating, cooling, and shaking, we use the QInstruments BioShake, which you can see in the back of the image.

Lastly, we included a tip comb change station, which I’ll explain next.

---

Specialty Consumables

To use the magnetic rod head, we need a specific tip comb and plate combination. The plate’s unique well geometry is not available off-the-shelf and required a custom adapter for the QInstruments BioShake.

Huge shoutout to our service engineering team—especially Salama—for designing and testing this adapter. We went through:

• Fit testing

• Thermal profiling

• Movement validation

After finalizing the adapter, we addressed the tip comb design. For magnetic rod extraction, the magnet enters the tip comb, which holds the magnetic beads (and thus the DNA) and moves them through washes and reagents. Each sample requires a unique tip comb.

Initially, we had no way to dynamically change tip combs. Each plate had its own comb, and the run couldn’t reuse positions, which was limiting on the LM1800.

---

Innovation: Tip Comb Exchange

Dynamic Devices worked closely with us on several solutions:

• Version 1: Manually placing four tip combs on the deck. It worked, but used too much space—12 positions for four tip combs. Not scalable.

• Version 2: The iMagZ, a pneumatic elevator that automates tip comb exchange. It elevates the used tip comb for disposal and then retracts for loading the next one. This solution is:

  • Space-efficient

  • Fully integrated

  • Currently in use in our production lab

While it’s mostly manual movements (e.g., not fully modular), it’s functional and works well for now.

---

Wrapping Up

We’ve seen major improvements from this work in terms of throughput, integration, and ease of operation. There’s still room for refinement—especially in recovery processes and automation robustness—but overall, the new platform has been a success.