We’re excited to share ChainSpace™, a framework we developed to bring sequence awareness into polymer design. In biology, sequence defines function—but in polymer science, sequence is often overlooked in favor of simpler metrics like monomer ratios. ChainSpace changes that by helping us explore the space of synthetically accessible sequences and how they affect material performance.

Why Sequence Is Often Ignored—and Why That’s a Problem

Most machine learning efforts in polymer design focus on what’s easy to control: which monomers to use and in what ratio. But that misses a huge part of the picture. Two polymers with the same composition can behave very differently depending on how those monomers are arranged. That’s the gap we built ChainSpace to address.

Four Ways to Control Polymer Sequence
ChainSpace organizes sequence design around four control levers:

• Compositional: choosing monomers and their ratios
• Kinetic: using reactivity differences to bias sequence
• Temporal: changing monomer feeds during synthesis
• Spatial: modifying polymers after synthesis using steric constraints

Each strategy opens access to different regions of sequence space, helping us design more precisely.

Describing Sequence with the Right Metrics

To make sense of sequence patterns, we use three statistical descriptors:

• λ₂ (local correlation): how blocky or alternating the sequence is
• Shannon entropy: how diverse the patterns are
• Mean run length: how long similar monomers cluster together

These give us a shared language to compare polymers, regardless of how they were made.

How This Helps in Gene Delivery

We applied ChainSpace to a set of polymers from our gene delivery platform and found clear links between sequence and performance. One of our top candidates had a moderately alternating pattern and short charged blocks—and hit 85% transfection efficiency. That kind of insight would be hard to get from composition alone.

Beyond Biology: Broader Applications in Materials Science

While we’ve focused on gene delivery, the underlying principles of ChainSpace extend well beyond biotech. Sequence-defined polymers play key roles in areas like:

• Semiconductors, where chain regularity affects charge transport
• Battery binders, where sequence can impact ion diffusion and interface stability
• Membranes and coatings, where surface behavior depends on how blocks and motifs are arranged
• Adhesives and emulsifiers, where gradient or blocky sequences tune flow, adhesion, or phase behavior

These systems all benefit from a clearer understanding of how sequence patterns translate to performance—and ChainSpace offers a way to map, quantify, and explore that space systematically.

What’s Next

ChainSpace helps us define better design spaces—ones that reflect what can actually be made in the lab and what might actually work in practice. It’s already helping us accelerate our polymer discovery pipeline, and we believe it has broader utility across the materials science landscape.

🔬 Read the full preprint here: https://chemrxiv.org/engage/chemrxiv/article-details/6883a2ff728bf9025e3699d4