Proteins are essential to life, and understanding their structure can facilitate a mechanistic understanding of their function. Through an enormous experimental effort1–4, the structures of around 100,000 unique proteins have been determined5, but this represents a small fraction of the billions of known protein sequences6,7. Structural coverage is bottlenecked by the months to years of painstaking effort required to determine a single protein structure. Accurate computational approaches are needed to address this gap and to enable large-scale structural bioinformatics. Predicting the three-dimensional structure that a protein will adopt based solely on its amino acid sequence—the structure prediction component of the ‘protein folding problem’8—has been an important open research problem for more than 50 years9. Despite recent progress10–14, existing methods fall far short of atomic accuracy, especially when no homologous structure is available. Here we provide the first computational method that can regularly predict protein structures with atomic accuracy even in cases in which no similar structure is known. We validated an entirely redesigned version of our neural network-based model, AlphaFold, in the challenging 14th Critical Assessment of protein Structure Prediction (CASP14)15, demonstrating accuracy competitive with experimental structures in a majority of cases and greatly outperforming other methods. Underpinning the latest version of AlphaFold is a novel machine learning approach that incorporates physical and biological knowledge about protein structure, leveraging multi-sequence alignments, into the design of the deep learning algorithm. AlphaFold predicts protein structures with an accuracy competitive with experimental structures in the majority of cases using a novel deep learning architecture.

/pdf/highly-accurate-protein-structure-prediction-with-alphafold-2xjbzfwu3n.pdf

Highly accurate protein structure prediction with AlphaFold

We describe the development, current features, and some directions for future development of the Amber package of computer programs. This package evolved from a program that was constructed in the late 1970s to do Assisted Model Building with Energy Refinement, and now contains a group of programs embodying a number of powerful tools of modern computational chemistry, focused on molecular dynamics and free energy calculations of proteins, nucleic acids, and carbohydrates.

/pdf/the-amber-biomolecular-simulation-programs-e2nyq1ef44.pdf

The Amber biomolecular simulation programs

Molecular mechanics is powerful for its speed in atomistic simulations, but an accurate force field is required. The Amber ff99SB force field improved protein secondary structure balance and dynamics from earlier force fields like ff99, but weaknesses in side chain rotamer and backbone secondary structure preferences have been identified. Here, we performed a complete refit of all amino acid side chain dihedral parameters, which had been carried over from ff94. The training set of conformations included multidimensional dihedral scans designed to improve transferability of the parameters. Improvement in all amino acids was obtained as compared to ff99SB. Parameters were also generated for alternate protonation states of ionizable side chains. Average errors in relative energies of pairs of conformations were under 1.0 kcal/mol as compared to QM, reduced 35% from ff99SB. We also took the opportunity to make empirical adjustments to the protein backbone dihedral parameters as compared to ff99SB. Multiple sm...

/pdf/ff14sb-improving-the-accuracy-of-protein-side-chain-and-2f0ci0lvo5.pdf

ff14SB: Improving the Accuracy of Protein Side Chain and Backbone Parameters from ff99SB

The ff94 force field that is commonly associated with the Amber simulation package is one of the most widely used parameter sets for biomolecular simulation. After a decade of extensive use and testing, limitations in this force field, such as over-stabilization of alpha-helices, were reported by us and other researchers. This led to a number of attempts to improve these parameters, resulting in a variety of "Amber" force fields and significant difficulty in determining which should be used for a particular application. We show that several of these continue to suffer from inadequate balance between different secondary structure elements. In addition, the approach used in most of these studies neglected to account for the existence in Amber of two sets of backbone phi/psi dihedral terms. This led to parameter sets that provide unreasonable conformational preferences for glycine. We report here an effort to improve the phi/psi dihedral terms in the ff99 energy function. Dihedral term parameters are based on fitting the energies of multiple conformations of glycine and alanine tetrapeptides from high level ab initio quantum mechanical calculations. The new parameters for backbone dihedrals replace those in the existing ff99 force field. This parameter set, which we denote ff99SB, achieves a better balance of secondary structure elements as judged by improved distribution of backbone dihedrals for glycine and alanine with respect to PDB survey data. It also accomplishes improved agreement with published experimental data for conformational preferences of short alanine peptides and better accord with experimental NMR relaxation data of test protein systems.

/pdf/comparison-of-multiple-amber-force-fields-and-development-of-2mpeu0hj18.pdf

Comparison of multiple Amber force fields and development of improved protein backbone parameters.

Recent advances in hardware and software have enabled increasingly long molecular dynamics (MD) simulations of biomolecules, exposing certain limitations in the accuracy of the force fields used for such simulations and spurring efforts to refine these force fields. Recent modifications to the Amber and CHARMM protein force fields, for example, have improved the backbone torsion potentials, remedying deficiencies in earlier versions. Here, we further advance simulation accuracy by improving the amino acid side-chain torsion potentials of the Amber ff99SB force field. First, we used simulations of model alpha-helical systems to identify the four residue types whose rotamer distribution differed the most from expectations based on Protein Data Bank statistics. Second, we optimized the side-chain torsion potentials of these residues to match new, high-level quantum-mechanical calculations. Finally, we used microsecond-timescale MD simulations in explicit solvent to validate the resulting force field against a large set of experimental NMR measurements that directly probe side-chain conformations. The new force field, which we have termed Amber ff99SB-ILDN, exhibits considerably better agreement with the NMR data. Proteins 2010. © 2010 Wiley-Liss, Inc.

https://www.onlinelibrary.wiley.com/doi/pdf/10.1002/prot.22711

Improved side‐chain torsion potentials for the Amber ff99SB protein force field

We present results from all-atom, fully unrestrained ab initio folding simulations for a stable protein with nontrivial secondary structure elements and a hydrophobic core. The construct, “trpcage”, is a 20-residue sequence optimized by the Andersen group at University of Washington and is currently the smallest protein that displays two-state folding properties. Compared over the well-defined regions of the experimental structure, our prediction has a remarkably low 0.97 A Cα root-mean-square-deviation (rmsd) and 1.4 A for all heavy atoms. The simulated structure family displays additional features that are suggested by experimental data, yet are not evident in the family of NMR-derived structures.

All-Atom Structure Prediction and Folding Simulations of a Stable Protein

The transferability of molecular mechanics parameters derived for small model systems to larger biopolymers such as proteins can be difficult to assess. Even for small peptides, molecular dynamics simulations are typically too short to sample structures significantly different than initial conformations, making comparison to experimental data questionable. We employed a PC cluster to generate large numbers of native and non-native conformations for peptides with experimentally measured structural data, one predominantly helical and the other forming a β-hairpin. These atomic-detail sets do not suffer from slow convergence, and can be used to rapidly evaluate important force field properties. In this case a suspected bias toward α-helical conformations in the ff94 and ff99 force fields distributed with the AMBER package was verified. The sets provide critical feedback not only on force field transferability, but may also predict modifications for improvement. Such predictions were used to modify the ff99 parameter set, and the resulting force field was used to test stability and folding of model peptides. Structural behavior during molecular dynamics with the modified force field is found to be very similar to expectations, suggesting that these basis sets of conformations may themselves have significant transferability among force fields. We continue to improve and expand this data set and plan to make it publicly accessible. The calculations involved in this process are trivially parallel and can be performed using inexpensive personal computers with commodity components. © 2002 Wiley Periodicals, Inc. J Comput Chem 24: 21–31, 2003

Using PC clusters to evaluate the transferability of molecular mechanics force fields for proteins.

Peptides based on C-terminal regions of the human immunodeficiency virus (HIV) viral protein gp41 represent an important new class of antiviral therapeutics called peptide fusion inhibitors. In this study, computational methods were used to model the binding of six peptides that contain residues that pack into a conserved hydrophobic pocket on HIVgp41, an attractive target site for the development of small molecule inhibitors. Free energies of binding were computed using molecular mechanics Generalized Born surface area (MM-GBSA) methods from molecular dynamics (MD) simulations, which employed either explicit (TIP3P) or continuum Generalized Born (GB) water models and strong correlations between experimental and computational affinities were obtained in both cases. Energy decomposition of the TIP3P-MD results (r2 = 0.75) reveals that variation in experimental affinity is highly correlated with changes in intermolecular van der Waals energies (ΔEvdw) on both a local (residue-based, r2 = 0.94) and global (peptide-based, r2 = 0.84) scale. The results show that differential association of C-peptides with HIVgp41 is driven solely by changes within the conserved pocket supporting the hypothesis that this region is an important drug target site. Such strong agreement with experiment is notable given the large size of the ligands (34 amino-acids) relative to the small range of experimental affinities (2 kcal/mol) and demonstrates good sensitivity of this computational method for simulating peptide fusion inhibitors. Finally, inspection of simulation trajectories identified a highly populated π-type hydrogen bond, which formed between Gln575 on the receptor and the aromatic ring of peptide ligand Phe631, which could have important implications for drug design. Proteins 2007. © 2007 Wiley-Liss, Inc.

Bentley Strockbine

Papers

Comparison of multiple Amber force fields and development of improved protein backbone parameters.

All-Atom Structure Prediction and Folding Simulations of a Stable Protein

Using PC clusters to evaluate the transferability of molecular mechanics force fields for proteins.

Binding of antifusion peptides with HIVgp41 from molecular dynamics simulations: Quantitative correlation with experiment