Directional ΔG Neural Network (DrΔG-Net): A Modular Neural Network Approach to Binding Free Energy Prediction.

The protein-ligand binding free energy is a central quantity in structure-based computational drug discovery efforts. Although popular alchemical methods provide sound statistical means of computing the binding free energy of a large breadth of systems, they are generally too costly to be applied at the same frequency as end point or ligand-based methods. By contrast, these data-driven approaches are typically fast enough to address thousands of systems but with reduced transferability to unseen systems. We introduce DrΔG-Net (or simply Dragnet), an equivariant graph neural network that can blend ligand-based and protein-ligand data-driven approaches. It is based on a 3D fingerprint representation of the ligand alone and in complex with the protein target. Dragnet is a global scoring function to predict the binding affinity of arbitrary protein-ligand complexes, but can be easily tuned via transfer learning to specific systems or end points, performing similarly to common 2D ligand-based approaches in these tasks. Dragnet is evaluated on a total of 28 validation proteins with a set of congeneric ligands derived from the Binding DB and one custom set extracted from the ChEMBL Database. In general, a handful of experimental binding affinities are sufficient to optimize the scoring function for a particular protein and ligand scaffold. When not available, predictions from physics-based methods such as absolute free energy perturbation can be used for the transfer learning tuning of Dragnet. Furthermore, we use our data to illustrate the present limitations of data-driven modeling of binding free energy predictions.

Tensor Hypercontraction Form of the Perturbative Triples Energy in Coupled-Cluster Theory.

We present the working equations for a reduced-scaling method of evaluating the perturbative triples (T) energy in coupled-cluster theory, through the tensor hypercontraction (THC) of the triples amplitudes (tijkabc). Through our method, we can reduce the scaling of the (T) energy from the traditional O(N7) to a more modest O(N5). We also discuss implementation details to aid future research, development, and software realization of this method. Additionally, we show that this method yields submillihartree (mEh) differences from CCSD(T) when evaluating absolute energies and sub-0.1 kcal/mol energy differences when evaluating relative energies. Finally, we demonstrate that this method converges to the true CCSD(T) energy through the systematic increasing of the rank or eigenvalue tolerance of the orthogonal projector, as well as exhibiting sublinear to linear error growth with respect to system size.

Out-of-hospital management of unresponsive, apneic, witnessed opioid overdoses: a case series from a supervised consumption site.

OBJECTIVES: There are conflicting recommendations for lay rescuer management of patients who are unresponsive and apneic due to opioid overdose. We evaluated the management of such patients at an urban supervised consumption site. METHODS: At a single urban supervised consumption site in Vancouver, BC, we conducted a retrospective chart review and administrative database linkage of consecutive patients who were unresponsive and apneic following witnessed opioid overdose between January 1, 2012 and December 31, 2017. We linked these visits with regional hospital records to define the entire care episode, which concluded when the patient was discharged from the supervised consumption site, ED, or hospital, or died. The primary outcome was successful resuscitation, defined as alive and neurologically intact (ambulatory and speaking coherently, or alert and oriented, or Glasgow Coma Scale 15) at the conclusion of the care episode. Secondary outcomes included mortality and predefined complications of resuscitation. RESULTS: We collected 767 patients, with a median age of 43 and 81.6% male, with complete follow-up on 763 patients (99.5%). All patients were managed with oxygen and ventilation (100%, 95% CI 0.995-1.0); 715 (93.2%, 95% CI 0.911-0.949) received naloxone; no patients underwent chest compressions (0%, 95% CI 0-0.005). All patients with complete follow-up were alive and neurologically intact at the end of their care episode (100%, 95% CI 0.994-1.0). Overall, 191 (24.9%) patients were transported to hospital, and 15 (2.0%) patients required additional naloxone after leaving the supervised consumption site; 16 (2.1%) developed complications, and 1 patient was admitted to hospital. CONCLUSIONS: At an urban supervised consumption site, all unresponsive, apneic patients with witnessed opioid overdose were successfully resuscitated with oxygen and/or naloxone. No patients required chest compressions.

RéSUMé: OBJECTIFS: Il existe des recommandations contradictoires concernant la prise en charge par des secouristes non professionnels des patients qui ne réagissent pas et sont apnéiques en raison d'une surdose d'opioïdes. Nous avons évalué la prise en charge de ces patients dans un site urbain de consommation supervisée. MéTHODES: Dans un seul site de consommation supervisée urbain à Vancouver, en Colombie-Britannique, nous avons effectué un examen rétrospectif des dossiers et un couplage de bases de données administratives de patients consécutifs qui étaient insensibles et apnéiques après avoir été témoins d'une surdose d'opioïdes entre le 1er janvier 2012 et le 31 décembre 2017. Le résultat primaire était la réussite de la réanimation, définie comme étant vivante et neurologiquement intacte (ambulatoire et parlant de manière cohérente, ou alerte et orientée, ou échelle de coma de Glasgow 15) à la fin de l'épisode de soins. Les résultats secondaires comprenaient la mortalité et les complications prédéfinies de la réanimation. RéSULTATS: Nous avons recueilli 767 patients, avec un âge médian de 43 ans et 81,6 % d'hommes, avec un suivi complet de 763 patients (99,5 %). Tous les patients ont été pris en charge avec de l'oxygène et la ventilation (100 %, IC à 95 % : 0,995-1,0) ; 715 (93,2 %, IC à 95 % : 0,911-0,949) ont reçu de la naloxone ; aucun patient n'a subi de compressions thoraciques (0 %, IC à 95 % : 0-0,005). Tous les patients ayant fait l'objet d'un suivi complet étaient vivants et intacts sur le plan neurologique à la fin de leur épisode de soins (100 %, IC à 95 % : 0,994-1,0). Dans l'ensemble, 191 (24,9 %) patients ont été transportés à l'hôpital, et 15 (2,0 %) patients ont eu besoin de naloxone supplémentaire après avoir quitté le site de consommation supervisée ; 16 (2,1 %) ont développé des complications, et 1 patient a été admis à l'hôpital. CONCLUSIONS: Dans un centre de consommation supervisée urbain, tous les patients apnéiques non réceptifs ayant été témoins d'une surdose d'opioïdes ont été réanimés avec succès avec de l'oxygène et/ou de la naloxone. Aucun patient n'a eu besoin de compressions thoraciques.

Comparison of rates of opioid withdrawal symptoms and reversal of opioid toxicity in patients treated with two naloxone dosing regimens: a retrospective cohort study.

INTRODUCTION: When managing opioid overdose (OD) patients, the optimal naloxone regimen should rapidly reverse respiratory depression while avoiding opioid withdrawal. Published naloxone administration guidelines have not been empirically validated and most were developed before fentanyl OD was common. In this study, rates of opioid withdrawal symptoms (OW) and reversal of opioid toxicity in patients treated with two naloxone dosing regimens were evaluated. METHODS: In this retrospective matched cohort study, health records of patients who experienced an opioid OD treated in two urban emergency departments (ED) during an ongoing fentanyl OD epidemic were reviewed. Definitions for OW and opioid reversal were developed a priori. Low dose naloxone (LDN; ≤0.15 mg) and high dose naloxone (HDN; >0.15 mg) patients were matched in a 1:4 ratio based upon initial respiratory rate (RR). The proportion of patients who developed OW and who met reversal criteria were compared between those treated initially with LDN or HDN. Odds ratios (OR) for OW and opioid reversal were obtained via logistic regression stratified by matched sets and adjusted for age, sex, pre-naloxone GCS, and presence of non-opioid drugs or alcohol. RESULTS: Eighty LDN patients were matched with 299 HDN patients. After adjustment, HDN patients were more likely than LDN patients to have OW after initial dose (OR = 8.43; 95%CI: 1.96, 36.3; p = 0.004) and after any dose (OR = 2.56; 95%CI: 1.17, 5.60; p = 0.019). HDN patients were more likely to meet reversal criteria after initial dose (OR = 2.73; 95%CI: 1.19, 6.26; p = 0.018) and after any dose (OR = 6.07; 95%CI: 1.81, 20.3; p = 0.003). CONCLUSIONS: HDN patients were more likely to have OW but also more likely to meet reversal criteria versus LDN patients.

Electron-Passing Neural Networks for Atomic Charge Prediction in Systems with Arbitrary Molecular Charge.

Atomic charges are critical quantities in molecular mechanics and molecular dynamics, but obtaining these quantities requires heuristic choices based on atom typing or relatively expensive quantum mechanical computations to generate a density to be partitioned. Most machine learning efforts in this domain ignore total molecular charges, relying on overfitting and arbitrary rescaling in order to match the total system charge. Here, we introduce the electron-passing neural network (EPNN), a fast, accurate neural network atomic charge partitioning model that conserves total molecular charge by construction. EPNNs predict atomic charges very similar to those obtained by partitioning quantum mechanical densities but at such a small fraction of the cost that they can be easily computed for large biomolecules. Charges from this method may be used directly for molecular mechanics, as features for cheminformatics, or as input to any neural network potential.

Inequities in Pediatric Abusive Head Trauma According to Neighborhood Social and Material Deprivation: A Population-Level Study in British Columbia, Canada.

OBJECTIVES: To explore the relationship between neighborhood social and material deprivation and the rates of abusive head trauma (AHT), and whether it differs according to sex, and following the implementation of the Period of PURPLE Crying (PURPLE) program. METHOD: A cross-sectional study design was applied to data from children 0 to 24 months old with a confirmed AHT diagnosis between 2005 and 2017 in British Columbia. Dissemination area-based social and material deprivation scores were assigned to residential areas, where AHT cases were recorded. Poisson regression models tested the relationship between deprivation scores and AHT rates, adding sex and pre-post program implementation as interaction terms. RESULTS: With each increase in material and social deprivation quintiles, AHT rates increased by 42% (95% CI [1.18, 1.72]) and 25% (95% CI [1.06, 1.51]), respectively, following a social gradient. AHT rate disparities between neighborhoods did not change following the PURPLE program implementation. CONCLUSIONS: This study stresses the need to provide additional AHT prevention services proportionately to the levels of neighborhood disadvantage, in addition to universal AHT programs, to successfully protect all children.