Best practices for constructing, preparing, and evaluating protein-ligand binding affinity benchmarks

Abstract

Free energy calculations are rapidly becoming indispensable in structure-enabled drug discovery programs. As new methods, force fields, and implementations are developed, assessing their expected accuracy on real-world systems (benchmarking) becomes critical to provide users with an assessment of the accuracy expected when these methods are applied within their domain of applicability, and developers with a way to assess the expected impact of new methodologies. These assessments require construction of a benchmark-a set of well-prepared, high quality systems with corresponding experimental measurements designed to ensure the resulting calculations provide a realistic assessment of expected performance when these methods are deployed within their domains of applicability. To date, the community has not yet adopted a common standardized benchmark, and existing benchmark reports suffer from a myriad of issues, including poor data quality, limited statistical power, and statistically deficient analyses, all of which can conspire to produce benchmarks that are poorly predictive of real-world performance. Here, we address these issues by presenting guidelines for (1) curating experimental data to develop meaningful benchmark sets, (2) preparing benchmark inputs according to best practices to facilitate widespread adoption, and (3) analysis of the resulting predictions to enable statistically meaningful comparisons among methods and force fields. We highlight challenges and open questions that remain to be solved in these areas, as well as recommendations for the collection of new datasets that might optimally serve to measure progress as methods become systematically more reliable. Finally, we provide a curated, versioned, open, standardized benchmark set adherent to these standards (protein-ligand-benchmark) and an open source toolkit for implementing standardized best practices assessments (arsenic) for the community to use as a standardized assessment tool. While our main focus is free energy methods based on molecular simulations, these guidelines should prove useful for assessment of the rapidly growing field of machine learning methods for affinity prediction as well.

Development and Benchmarking of Open Force Field v1.0.0—the Parsley Small-Molecule Force Field

“…We assessed the performance of the newly fitted Parsley force field in binding free energy calculations based on molecular dynamics simulations following suggested best practices for benchmarking binding affinities. 81 The test set consisted of eight protein targets with a total of 199 ligands (Supporting Information, Table S2), using a set commonly referred to as the "JACS data set" from a prior study published in that journal and frequently used by the community. 12 For a fair comparison to previously published results, the initial ligand and protein structures were obtained from prior benchmark work.…”

Assessing the effect of forcefield parameter sets on the accuracy of relative binding free energy calculations

“…It is worth noting that experimental uncertainties can be on the order of 0.64 kcal/mol. ( Hahn et al, 2021 ). Starting with a parameter set using the GAFF 2.11 ligand force field, the AMBER ff14SB protein forcefield, the SPC/E water and AM1-BCC charges, the overall mean unsigned error (MUE) and root mean square error (RMSE) of 199 ligands were 0.89 kcal/mol and 1.15 kcal/mol, respectively ( Table 2 ).…”

“…However, concerns were raised with the quality of the JNK1 structure (2GMX) used for benchmarking. ( Hahn et al, 2021 ). The high R-free value (0.351), as well as the large difference between R-value and R-free for this JNK1 structure, indicates a possible overfit of the atomic model to the experimental diffraction pattern when solving the crystal structure.…”

“…Although the better performance was seen in the case of JNK1, the protein structure used (2GMX) does not fulfill the high-quality structure criteria. This issue flags the importance of adopting best practices in constructing, preparing, and evaluating FEP calculations ( Mey et al, 2020 ; Hahn et al, 2021 ). Finally, most of the poorly predicted compounds (MUE >2.0 kcal/mol) fall into three cases (BACE, MCL1 and PTP1B), where the ligands are charged.…”

Scoring Functions for Protein-Ligand Binding Affinity Prediction Using Structure-based Deep Learning: A Review

“…Rigorous computational methods based on all-atom molecular dynamics simulations in explicit solvent-such as free energy perturbation and thermodynamic integration (Adcock and McCammon 2006)-can compute accurate relative and absolute binding free energies (Bash et al, 1987;Boresch et al, 2003;Mobley et al, 2007;Aldeghi et al, 2016, Aldeghi et al, 2018aCournia et al, 2017), predict ligand selectivity (Aldeghi et al, 2017) and mutation effects (Aldeghi et al, 2018b;Hauser et al, 2018), and guide fragment elaborations (Alibay et al, 2022). Unfortunately, such rigorous methods are computationally expensive and often require a lot of expert knowledge and domain expertise (Mey et al, 2020;Hahn et al, 2021). This remains true even for simpler methods such as ligand-interaction energy (LIE) (Åqvist et al, 1994;Jones-Hertzog and Jorgensen 1997).…”