FiberForge: enabling high-throughput simulations of the mechanical properties of helical fibrils

Abstract

The mechanical properties of amyloid-based materials are governed by fibril geometry, sequence, and polymorphism, yet systematic exploration of this vast design space has been limited by the lack of high-throughput modeling tools. Here we present FiberForge, an open-source workflow that automates construction of amyloid protofibrils, streamlines high-throughput simulations of amyloid deformation, and analyzes fibril trajectories to estimate mechanical properties and fracture mechanisms. Using 374 full-length amyloid crystal structures from the Protein Data Bank, FiberForge rebuilds fibrils with a mean per-chain RMSD of 1.7 Å (median 2.2 Å), demonstrating accurate structural recovery across wide sequence (18–420 aa) and symmetry ranges. Extensive SMD benchmarking on Aβ(1–42) (2MXU) yields a mean rupture force of 1.534 ± 0.164 nN from 232 replicas; bootstraping analysis shows that three replicas suffice for converged elastic-modulus and strength estimates. High-throughput screening of the amyloid fiber dataset produces elastic moduli of 0.2–20 GPa and ultimate tensile strengths of 0.1–1 GPa. Comparison with four AFM-characterized systems shows agreement within an order of magnitude, underscoring the method's predictive capability. FiberForge's screening results also enable larger-scale sequence–structure–property analysis, revealing that mechanical behavior is correlated with molecular assembly geometry, especially hydrogen-bond density. While earlier work suggested the relevance of these features for particular systems, our results demonstrate their importance across diverse fibril architectures. FiberForge thus provides an end-to-end platform for molecular modeling and design of amyloid materials, enabling physics-based identification of sequences and polymorphs with targeted mechanical behavior.