FEP Protocol Builder: Optimization of Free Energy Perturbation Protocols Using Active Learning.

Significant improvements have been made in the past decade to methods that rapidly and accurately predict binding affinity through free energy perturbation (FEP) calculations. This has been driven by recent advances in small-molecule force fields and sampling algorithms combined with the availability of low-cost parallel computing. Predictive accuracies of ∼1 kcal mol-1 have been regularly achieved, which are sufficient to drive potency optimization in modern drug discovery campaigns. Despite the robustness of these FEP approaches across multiple target classes, there are invariably target systems that do not display expected performance with default FEP settings. Traditionally, these systems required labor-intensive manual protocol development to arrive at parameter settings that produce a predictive FEP model. Due to the (a) relatively large parameter space to be explored, (b) significant compute requirements, and (c) limited understanding of how combinations of parameters can affect FEP performance, manual FEP protocol optimization can take weeks to months to complete, and often does not involve rigorous train-test set splits, resulting in potential overfitting. These manual FEP protocol development timelines do not coincide with tight drug discovery project timelines, essentially preventing the use of FEP calculations for these target systems. Here, we describe an automated workflow termed FEP Protocol Builder (FEP-PB) to rapidly generate accurate FEP protocols for systems that do not perform well with default settings. FEP-PB uses an active-learning workflow to iteratively search the protocol parameter space to develop accurate FEP protocols. To validate this approach, we applied it to pharmaceutically relevant systems where default FEP settings could not produce predictive models. We demonstrate that FEP-PB can rapidly generate accurate FEP protocols for the previously challenging MCL1 system with limited human intervention. We also apply FEP-PB in a real-world drug discovery setting to generate an accurate FEP protocol for the p97 system. FEP-PB is able to generate a more accurate protocol than the expert user, rapidly validating p97 as amenable to free energy calculations. Additionally, through the active-learning workflow, we are able to gain insight into which parameters are most important for a given system. These results suggest that FEP-PB is a robust tool that can aid in rapidly developing accurate FEP protocols and increasing the number of targets that are amenable to the technology.

Noncovalent inhibitors reveal BTK gatekeeper and auto-inhibitory residues that control its transforming activity.

Inhibition of Bruton tyrosine kinase (BTK) is a breakthrough therapy for certain B cell lymphomas and B cell chronic lymphatic leukemia. Covalent BTK inhibitors (e.g., ibrutinib) bind to cysteine C481, and mutations of this residue confer clinical resistance. This has led to the development of noncovalent BTK inhibitors that do not require binding to cysteine C481. These new compounds are now entering clinical trials. In a systematic BTK mutagenesis screen, we identify residues that are critical for the activity of noncovalent inhibitors. These include a gatekeeper residue (T474) and mutations in the kinase domain. Strikingly, co-occurrence of gatekeeper and kinase domain lesions (L512M, E513G, F517L, L547P) in cis results in a 10- to 15-fold gain of BTK kinase activity and de novo transforming potential in vitro and in vivo. Computational BTK structure analyses reveal how these lesions disrupt an intramolecular mechanism that attenuates BTK activation. Our findings anticipate clinical resistance mechanisms to a new class of noncovalent BTK inhibitors and reveal intramolecular mechanisms that constrain BTK's transforming potential.

Application of Free Energy Perturbation (FEP+) to Understanding Ligand Selectivity: A Case Study to Assess Selectivity Between Pairs of Phosphodiesterases (PDE's).

Cyclic nucleotide phosphodiesterases (PDE's) are metalloenzymes that play a key role in regulating the levels of the ubiquitous second messengers, cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP). In humans, 11 PDE protein families mediate numerous biochemical pathways throughout the body and are effective drug targets for the treatment of diseases ranging from central nervous system disorders to heart and pulmonary diseases. PDE's also share a highly conserved catalytic site (about 50%), thus making the design of selective drug candidates very challenging with classical structure-based design approaches given also the lack of publicly available co-crystal structures of pairs of PDE's with consistent biological data to be compared, as we show in our work. In this retrospective study, we apply free energy perturbation (FEP+) to predict the selectivity of inhibitors that bind two pairs of closely related PDE families: PDE9/1 and PDE5/6 where only 1 co-crystal structure per pair is publicly available. As another challenge, the p Ka of the PDE5/6 inhibitor is close to the experimental pH, making unclear the exact protonation state that should be used in the computational workflow. We demonstrate that running FEP+ on homology models constructed for these metalloenzymes accurately reproduces experimentally observed selectivity profiles also addressing the unclear protonation state to be used during computation with our recently developed p Ka-correction method. Based on these data, we conclude that FEP+ is a robust method for prediction of selectivity for this target class and may be helpful to address related lead optimization challenges in drug discovery.

Predicting resistance of clinical Abl mutations to targeted kinase inhibitors using alchemical free-energy calculations.

The therapeutic effect of targeted kinase inhibitors can be significantly reduced by intrinsic or acquired resistance mutations that modulate the affinity of the drug for the kinase. In cancer, the majority of missense mutations are rare, making it difficult to predict their impact on inhibitor affinity. This complicates the practice of precision medicine, pairing of patients with clinical trials, and development of next-generation inhibitors. Here, we examine the potential for alchemical free-energy calculations to predict how kinase mutations modulate inhibitor affinities to Abl, a major target in chronic myelogenous leukemia (CML). We find these calculations can achieve useful accuracy in predicting resistance for a set of eight FDA-approved kinase inhibitors across 144 clinically-identified point mutations, achieving a root mean square error in binding free energy changes of 1.10.91.3 kcal/mol (95% confidence interval) and correctly classifying mutations as resistant or susceptible with 888293% accuracy. Since these calculations are fast on modern GPUs, this benchmark establishes the potential for physical modeling to collaboratively support the rapid assessment and anticipation of the potential for patient mutations to affect drug potency in clinical applications.

Accurate and Reliable Prediction of the Binding Affinities of Macrocycles to Their Protein Targets.

Macrocycles have been emerging as a very important drug class in the past few decades largely due to their expanded chemical diversity benefiting from advances in synthetic methods. Macrocyclization has been recognized as an effective way to restrict the conformational space of acyclic small molecule inhibitors with the hope of improving potency, selectivity, and metabolic stability. Because of their relatively larger size as compared to typical small molecule drugs and the complexity of the structures, efficient sampling of the accessible macrocycle conformational space and accurate prediction of their binding affinities to their target protein receptors poses a great challenge of central importance in computational macrocycle drug design. In this article, we present a novel method for relative binding free energy calculations between macrocycles with different ring sizes and between the macrocycles and their corresponding acyclic counterparts. We have applied the method to seven pharmaceutically interesting data sets taken from recent drug discovery projects including 33 macrocyclic ligands covering a diverse chemical space. The predicted binding free energies are in good agreement with experimental data with an overall root-mean-square error (RMSE) of 0.94 kcal/mol. This is to our knowledge the first time where the free energy of the macrocyclization of linear molecules has been directly calculated with rigorous physics-based free energy calculation methods, and we anticipate the outstanding accuracy demonstrated here across a broad range of target classes may have significant implications for macrocycle drug discovery.

OPLS3: A Force Field Providing Broad Coverage of Drug-like Small Molecules and Proteins.

The parametrization and validation of the OPLS3 force field for small molecules and proteins are reported. Enhancements with respect to the previous version (OPLS2.1) include the addition of off-atom charge sites to represent halogen bonding and aryl nitrogen lone pairs as well as a complete refit of peptide dihedral parameters to better model the native structure of proteins. To adequately cover medicinal chemical space, OPLS3 employs over an order of magnitude more reference data and associated parameter types relative to other commonly used small molecule force fields (e.g., MMFF and OPLS_2005). As a consequence, OPLS3 achieves a high level of accuracy across performance benchmarks that assess small molecule conformational propensities and solvation. The newly fitted peptide dihedrals lead to significant improvements in the representation of secondary structure elements in simulated peptides and native structure stability over a number of proteins. Together, the improvements made to both the small molecule and protein force field lead to a high level of accuracy in predicting protein-ligand binding measured over a wide range of targets and ligands (less than 1 kcal/mol RMS error) representing a 30% improvement over earlier variants of the OPLS force field.

Predicting Binding Affinities for GPCR Ligands Using Free-Energy Perturbation.

The rapid growth of structural information for G-protein-coupled receptors (GPCRs) has led to a greater understanding of their structure, function, selectivity, and ligand binding. Although novel ligands have been identified using methods such as virtual screening, computationally driven lead optimization has been possible only in isolated cases because of challenges associated with predicting binding free energies for related compounds. Here, we provide a systematic characterization of the performance of free-energy perturbation (FEP) calculations to predict relative binding free energies of congeneric ligands binding to GPCR targets using a consistent protocol and no adjustable parameters. Using the FEP+ package, first we validated the protocol, which includes a full lipid bilayer and explicit solvent, by predicting the binding affinity for a total of 45 different ligands across four different GPCRs (adenosine A2AAR, ß1 adrenergic, CXCR4 chemokine, and δ opioid receptors). Comparison with experimental binding affinity measurements revealed a highly predictive ranking correlation (average spearman ρ = 0.55) and low root-mean-square error (0.80 kcal/mol). Next, we applied FEP+ in a prospective project, where we predicted the affinity of novel, potent adenosine A2A receptor (A2AR) antagonists. Four novel compounds were synthesized and tested in a radioligand displacement assay, yielding affinity values in the nanomolar range. The affinity of two out of the four novel ligands (plus three previously reported compounds) was correctly predicted (within 1 kcal/mol), including one compound with approximately a tenfold increase in affinity compared to the starting compound. Detailed analyses of the simulations underlying the predictions provided insights into the structural basis for the two cases where the affinity was overpredicted. Taken together, these results establish a protocol for systematically applying FEP+ to GPCRs and provide guidelines for identifying potent molecules in drug discovery lead optimization projects.

A combination of intrathecal morphine and remifentanil anesthesia for fast-track cardiac anesthesia and surgery.

OBJECTIVE: To determine if the combined remifentanil and intrathecal morphine (RITM) anesthetic technique facilitates early extubation in patients undergoing coronary artery bypass graft (CABG) surgery. DESIGN: Prospective, randomized, controlled clinical trial. SETTING: Referral center for cardiothoracic surgery at a university hospital. PARTICIPANTS: Patients (n = 24) undergoing first-time elective CABG surgery. INTERVENTIONS: Two groups represented RITM (n = 12) and fentanyl-based (controls, n = 12) anesthesia. Premedication was standardized to temazepam, 0.4 mg/kg, and anesthesia was induced with etomidate, 0.3 mg/kg, in both groups. The RITM group received remifentanil, 1 microg/kg bolus followed by 0.25 to 1 microg/kg/min infusion, and intrathecal morphine, 2 mg. The control group received fentanyl, 12 microg/kg in 3 divided doses. Anesthesia was maintained with isoflurane and pancuronium in both groups. After completion of surgery, the remifentanil infusion was stopped. Complete reversal of muscle relaxation was ensured with a nerve stimulator, and a propofol infusion, 0.5 to 3 mg/kg/h, was started in both groups. All patients were transferred to the intensive care unit (ICU) to receive standardized postoperative care. Intensivists and ICU nurses were blinded to the group assignment. Propofol infusion was stopped, and the tracheal extubation was accomplished when extubation criteria were fulfilled. MEASUREMENTS AND MAIN RESULTS: Both groups were similar with respect to demographic data and surgical characteristics. Extubation times were 156 +/- 82 minutes and 258 +/- 91 minutes in the RITM and control groups (p = 0.012). Patients in the RITM group exhibited lower visual analog scale pain scores during the first 2 hours after extubation (p < 0.04). Morphine requirements during the 24 hours after extubation were 2.5 +/- 3 mg in the RITM group and 16 +/- 11 mg in the control group (p = 0.0018). Sedation scores were lower in the RITM group during the first 3 hours after extubation (p < 0.03). Pulmonary function tests as assessed by spirometry were better in the RITM group at 6 and 12 hours after extubation (p < 0.04). There were no significant differences in PaO(2) and PaCO(2) after extubation between the 2 groups. None of the patients had episodes of apnea during the immediate 24-hour postextubation period. Two patients from the RITM group required reintubation on the second and sixth postoperative days. There were no differences in ICU and hospital length of stay between the 2 groups. CONCLUSION: Implementation of the RITM technique provided earlier tracheal extubation, decreased level of sedation, excellent analgesia, and improved spirometry in the early postoperative period. The impact of RITM on ICU and hospital length of stay and potential cost benefits require further evaluation.