Elfin: An algorithm for the computational design of custom three-dimensional structures from modular repeat protein building blocks

Abstract

Computational protein design methods have enabled the design of novel protein structures, but they are often still limited to small proteins and symmetric systems. To expand the size of designable proteins while controlling the overall structure, we developed Elfin, a genetic algorithm for the design of novel proteins with custom shapes using structural building blocks derived from experimentally verified repeat proteins. By combining building blocks with compatible interfaces, it is possible to rapidly build non-symmetric large structures (>1000 amino acids) that match three-dimensional geometric descriptions provided by the user. A run time of about 20min on a laptop computer for a 3000 amino acid structure makes Elfin accessible to users with limited computational resources. Protein structures with controlled geometry will allow the systematic study of the effect of spatial arrangement of enzymes and signaling molecules, and provide new scaffolds for functional nanomaterials.

Elfin UI: A Graphical Interface for Protein Design With Modular Building Blocks

“…The Elfin software package is built around the Elfin solver, a genetic algorithm for the assembly of modular structures matching a user defined shape (Yeh et al, 2018), and contains an updated database with information about modular building blocks, a graphical user interface (Elfin UI) built as Blender add-on, and ancillary utility scripts (e.g., for installation, database preparation, file conversion). Code, documentation, installation scripts and tutorials are available on https://github.…”

“…Protein design tools focus largely on atomic models and sequence design from a given backbone structure. Additionally, several approaches allow to build completely new structures by relying on secondary structure description and fragments assembly, like Rosetta remodel and blueprint builder (Huang et al, 2011;Koga et al, 2012), parametric design, as in Isambard (Wood et al, 2017), idealized secondary structures, e.g., CoCoPOD (Ljubetič et al, 2017) and TopoBuilder (Sesterhenn et al, 2020), or building blocks with super-secondary structures, as in SEWING (Jacobs et al, 2016) and Elfin (Yeh et al, 2018). Protein complexes have been successfully designed for symmetric systems, e.g., point group symmetry (Lai et al, 2012;King et al, 2014;Hsia et al, 2016) and lattices (Lanci et al, 2012;Gonen et al, 2015), but large, precise and asymmetric assemblies are still a challenge.…”

“…In this work we have developed a user interface to allow design of protein structures using modular structural building blocks. Elfin user interface (Elfin UI) was developed as a graphical interface and an interactive editor to the Elfin software package (Yeh et al, 2018) for design of custom protein architectures (Figure 1). Elfin uses structural compatible building blocks (referred to as modules) derived from experimentally validated structures of repeat proteins to build large and complex architectures.…”

Repeat proteins: designing new shapes and functions for solenoid folds

“…Elfin [21] and its graphical user interface Elfin UI [22] have been developed for the design of protein architectures using existing building blocks with known interfaces. Repeat protein units and helical junctions can be combined directly, if interfaces are compatible (e.g.…”

Combining statistical and neural network approaches to derive energy functions for completely flexible protein backbone design

“…On the other hand, current methods for designing protein backbones still heavily rely on structure type-specific heuristic rules or parametric models. (Grigoryan and Degrado, 2011;Huang, et al, 2014;Jacobs, et al, 2016;Lin, et al, 2015;Yeh, et al, 2018) To take full advantage of the plasticity of protein backbone conformations in protein design, (Huang, et al, 2016;MacDonald and Freemont, 2016) it is highly desirable to have a general method that can be used to sample and optimize designable polypeptide backbones without pre-specified amino acid sequences. While a few previous computational studies have suggested that well-folded protein conformations may correspond to minima on free energy surfaces of backbones modeled without specific sidechain information, (Cossio, et al, 2010;Hoang, et al, 2004;Kukic, et al, 2015;Taylor, et al, 2009;Zhang, et al, 2006) most of these studies have aimed at coarsely contouring the free energy landscapes of polypeptides rather than at obtaining accurate backbone structures to be used for amino acid sequence design.…”