Protein design with a comprehensive statistical energy function and boosted by experimental selection for foldability

Abstract

The de novo design of amino acid sequences to fold into desired structures is a way to reach a more thorough understanding of how amino acid sequences encode protein structures and to supply methods for protein engineering. Notwithstanding significant breakthroughs, there are noteworthy limitations in current computational protein design. To overcome them needs computational models to complement current ones and experimental tools to provide extensive feedbacks to theory. Here we develop a comprehensive statistical energy function for protein design with a new general strategy and verify that it can complement and rival current well-established models. We establish that an experimental approach can be used to efficiently assess or improve the foldability of designed proteins. We report four de novo proteins for different targets, all experimentally verified to be well-folded, solved solution structures for two being in excellent agreement with respective design targets.

On statistical energy functions for biomolecular modeling and design

“…Estimating the one-residue and two-residue energies Next, we assume that the total energy can be broken into single body or one-residue terms, or the sum of singlebody and two-body or two-residue terms, ignoring the remaining higher order interactions [10,16,43]. This leads Figure 3.…”

“…In the current paper, we try to analyze the topic with a compact and theoretically-oriented point of view. While some of the issues discussed here should have already been examined in existing literature, albeit to varied extents [16,[29][30][31][32][33], previous discussions have been widely scattered at different places and often in disparate contexts. In this review, the various aspects are introduced and discussed in a related context.…”

“…Instead, we will focus on statistical energy functions (also called knowledge-based energy functions) that are meant to be generally applicable to different structural types or protein families. Such statistical energy functions can be used alone, or they can be used in combination with physicsbased models of complementing strengths [18,19], to boost the accuracy and efficiency of both structural modeling and sequence design of biomolecules, especially proteins [16,20].…”

“…The statistical energy functions for sequence design are closely related to those for structural modeling, but involve some distinct considerations. Resolving some of these issues has led us to develop the ABACUS energy function, which has been shown to complement and rival previous best methods in fixed-backbone de novo protein design [16], with experimentally verified design results. ABACUS is acronym for A Backbone-based Amino aCid Usage Survey.…”

“…At the foundations of these approaches are computational models developed to quantify the interand intra-molecular interactions between individual atoms or residues. These include physics-based models such as the various molecular mechanics force fields [5][6][7], and data-based models such as the so-called statistical or knowledge-based potential energy functions [8][9][10][11][12][13][14][15][16].…”