De novo design of peptides localizing at the interface of biomolecular condensates

2026-05-16 · Source: Machine learning : nature.com subject feeds · Field: Science & Research — Life Sciences & Biology, Mathematics & Computational Sciences, Engineering & Applied Sciences · Depth: Expert, quick

Summary

Researchers at ETH Zurich developed a computational pipeline for the de novo design of peptides that localize at the interfaces of biomolecular condensates. This pipeline integrates high-throughput coarse-grained simulations, machine learning, and mixed-integer linear programming. The team validated their workflow by designing and synthesizing peptides that successfully localized at the interface of three distinct condensates formed by intrinsically disordered protein regions. These designed peptides exhibited surfactant-like architectures, featuring one tail enriched in aromatic residues that inserts into the condensate, and an opposite tail excluded from the dense phase, with its sequence varying based on the scaffold's net charge. This strategy offers a general approach for rationally designing interface-localizing peptides and elucidating underlying design principles.

Key takeaway

For AI Scientists and Research Scientists focused on biomolecular engineering, this pipeline offers a robust method to rationally design peptides for specific condensate interfaces. You can apply this computational strategy to engineer novel biomolecular systems or to further unravel the molecular grammar governing condensate localization, potentially impacting drug delivery or synthetic biology applications.

Key insights

A computational pipeline enables de novo design of peptides targeting biomolecular condensate interfaces.

Principles

• Interface localization is driven by surfactant-like peptide architectures.
• Aromatic residues enrich the condensate-inserting peptide tail.

Method

The pipeline combines high-throughput coarse-grained simulations, machine learning, and mixed-integer linear programming to design peptides.

In practice

• Design peptides for specific condensate targets.
• Engineer condensate interfaces for protein aggregation control.

Topics

• De Novo Peptide Design
• Biomolecular Condensates
• Computational Pipeline
• Coarse-Grained Simulations
• Machine Learning

Best for: AI Scientist, Research Scientist

Related on AIssential

• De novo molecular generation with optical property preconditioning at the token level — Token-conditioned autoregressive models can generate OLED molecules with targeted optical properties, but control varies by chemical motif.
• All life runs on 20 amino acids. These cells run key machinery on just 19 — Bacteria were reengineered to operate with 19 amino acids, leveraging AI for protein design.
• MultiMolecule: a modular ecosystem for biomolecular sequence-model workflows — MultiMolecule standardizes biomolecular sequence model reuse by providing complete, source-checked implementations with shared interfaces and provenance.
• From possibility to precision in macromolecular ensemble prediction — Predicting dynamic protein conformational ensembles requires integrating diverse experimental data and advanced AI.
• Bridging three-dimensional molecular structures and artificial intelligence with a conformation description language — ConfSeq enables large language models to process 3D molecular structures by converting them into discrete token sequences.
• GPT-5 lowers the cost of cell-free protein synthesis — GPT-5 connected to a robotic lab significantly reduced cell-free protein synthesis costs through autonomous closed-loop optimization.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine learning : nature.com subject feeds.