Dock2D: Synthetic data for the molecular recognition problem.

Predicting the physical interaction of proteins is a cornerstone problem in computational biology. New classes of learning-based algorithms are actively being developed, and are typically trained end-to-end on protein complex structures extracted from the Protein Data Bank. These training datasets tend to be large and difficult to use for prototyping and, unlike image or natural language datasets, they are not easily interpretable by non-experts. We present Dock2D-IP and Dock2DIF, two "toy" datasets that can be used to select algorithms predicting protein-protein interactions-or any other type of molecular interactions. Using two-dimensional shapes as input, each example from Dock2D-IP ("interaction pose") describe the interaction pose of two shapes known to interact and each example from Dock2D-IF ("interaction fact") describes whether two shapes form a stable complex or not, regardless of how they bind. We propose a number of baseline solutions to the problem and show that the same underlying energy function can be learned either by solving the interaction pose task (formulated as an energy-minimization "docking" problem) or the fact-ofinteraction task (formulated as a binding free energy estimation problem).

Statistical mechanics of polarizable force fields based on classical Drude oscillators with dynamical propagation by the dual-thermostat extended Lagrangian.

Polarizable force fields based on classical Drude oscillators offer a practical and computationally efficient avenue to carry out molecular dynamics (MD) simulations of large biomolecular systems. To treat the polarizable electronic degrees of freedom, the Drude model introduces a virtual charged particle that is attached to its parent nucleus via a harmonic spring. Traditionally, the need to relax the electronic degrees of freedom for each fixed set of nuclear coordinates is achieved by performing an iterative self-consistent field (SCF) calculation to satisfy a selected tolerance. This is a computationally demanding procedure that can increase the computational cost of MD simulations by nearly one order of magnitude. To avoid the costly SCF procedure, a small mass is assigned to the Drude particles, which are then propagated as dynamic variables during the simulations via a dual-thermostat extended Lagrangian algorithm. To help clarify the significance of the dual-thermostat extended Lagrangian propagation in the context of the polarizable force field based on classical Drude oscillators, the statistical mechanics of a dual-temperature canonical ensemble is formulated. The conditions for dynamically maintaining the dual-temperature properties in the case of the classical Drude oscillator are analyzed using the generalized Langevin equation.

Salt-Dependent Interactions between the C-Terminal Domain of Osmoregulatory Transporter ProP of Escherichia coli and the Lipid Membrane.

Osmosensing transporter ProP detects the increase in cytoplasmic cation concentration associated with osmotically induced cell dehydration and mediates osmolyte uptake into bacteria. ProP is a 12-transmembrane helix protein with an α-helical, cytoplasmic C-terminal domain (CTD) linked to transmembrane helix XII (TM XII). It has been proposed that the CTD helix associates with the anionic membrane surface to lock ProP in an inactive conformation and that the release of the CTD may activate ProP. To investigate this possible activation mechanism, we have built and simulated a structural model in which the CTD was anchored to the membrane by TM XII and the CTD helix was associated with the membrane surface. Molecular dynamics simulations showed specific intrapeptide salt bridges forming when the CTD associated with the membrane. Experiments supported the presence of the salt bridge Lys447-Asp455 and suggested a role for these residues in osmosensing. Simulations performed at different salt concentrations showed weakened CTD-lipid interactions at 0.25 M KCl and gradual stiffening of the membrane with increasing salinity. These results suggest that salt cations may affect CTD release and activate ProP by increasing the order of membrane phospholipids.

The Impact of Composition and Morphology on Ionic Conductivity of Silk/Cellulose Bio-Composites Fabricated from Ionic Liquid and Varying Percentages of Coagulation Agents.

Blended biocomposites created from the electrostatic and hydrophobic interactions between polysaccharides and structural proteins exhibit useful and unique properties. However, engineering these biopolymers into applicable forms is challenging due to the coupling of the material's physicochemical properties to its morphology, and the undertaking that comes with controlling this. In this particular study, numerous properties of the Bombyx mori silk and microcrystalline cellulose biocomposites blended using ionic liquid and regenerated with various coagulation agents were investigated. Specifically, the relationship between the composition of polysaccharide-protein bio-electrolyte membranes and the resulting morphology and ionic conductivity is explored using numerous characterization techniques, including scanning electron microscopy (SEM), Fourier transform infrared spectroscopy (FTIR), thermal gravimetric analysis (TGA), differential scanning calorimetry (DSC), X-ray scattering, atomic force microscopy (AFM) based nanoindentation, and dielectric relaxation spectroscopy (DRS). The results revealed that when silk is the dominating component in the biocomposite, the ionic conductivity is higher, which also correlates with higher ß-sheet content. However, when cellulose becomes the dominating component in the biocomposite, this relationship is not observed; instead, cellulose semicrystallinity and mechanical properties dominate the ionic conduction.

Drude polarizable force field for cation-π interactions of alkali and quaternary ammonium ions with aromatic amino acid side chains.

Cation-π interactions play important roles in molecular recognition and in the stability and function of proteins. However, accurate description of the structure and energetics of cation-π interactions presents a challenge to both additive and polarizable force fields, which are rarely designed to account for the complexation of charged groups with aromatic moieties. We calibrate the Drude polarizable force field for complexes of alkali metal ions (Li+ , Na+ , K+ , Rb+ , Cs+ ), ammonium (NH4+ ), tetramethylammonium (TMA+ ), and tetraethylammonium (TEA+ ) with aromatic amino acid side chain model compounds (benzene, toluene, 4-methylphenol, 3-methylindole) using high-level ab initio quantum chemical properties of these complexes. Molecular dynamics simulations reveal that cation-π complexes of the hard and tightly coordinated Li+ and Na+ ions are not stable in water but that larger ions form stable complexes, with binding free energies ranging between -0.8 and -2.9 kcal/mol. Like in gas phase, all complexes at equilibrium adopt an "en-face" complexation mode in water. The optimized Drude polarizable model provides an accurate description of the cation-π interactions involving small ions and proteins. © 2019 Wiley Periodicals, Inc.

Genetic, Structural, and Functional Evidence Link TMEM175 to Synucleinopathies.

OBJECTIVE: The TMEM175/GAK/DGKQ locus is the 3rd strongest risk locus in genome-wide association studies of Parkinson disease (PD). We aimed to identify the specific disease-associated variants in this locus, and their potential implications. METHODS: Full sequencing of TMEM175/GAK/DGKQ followed by genotyping of specific associated variants was performed in PD (n = 1,575) and rapid eye movement sleep behavior disorder (RBD) patients (n = 533) and in controls (n = 1,583). Adjusted regression models and a meta-analysis were performed. Association between variants and glucocerebrosidase (GCase) activity was analyzed in 715 individuals with available data. Homology modeling, molecular dynamics simulations, and lysosomal localization experiments were performed on TMEM175 variants to determine their potential effects on structure and function. RESULTS: Two coding variants, TMEM175 p.M393T (odds ratio [OR] = 1.37, p = 0.0003) and p.Q65P (OR = 0.72, p = 0.005), were associated with PD, and p.M393T was also associated with RBD (OR = 1.59, p = 0.001). TMEM175 p.M393T was associated with reduced GCase activity. Homology modeling and normal mode analysis demonstrated that TMEM175 p.M393T creates a polar side-chain in the hydrophobic core of the transmembrane, which could destabilize the domain and thus impair either its assembly, maturation, or trafficking. Molecular dynamics simulations demonstrated that the p.Q65P variant may increase stability and ion conductance of the transmembrane protein, and lysosomal localization was not affected by these variants. INTERPRETATION: Coding variants in TMEM175 are likely to be responsible for the association in the TMEM175/GAK/DGKQ locus, which could be mediated by affecting GCase activity. ANN NEUROL 2020;87:139-153.

Gold nanoparticles in ophthalmology.

Many research projects are underway to improve the diagnosis and therapy in ophthalmology. Indeed, visual acuity deficits affect 285 million people worldwide and different strategies are being developed to strengthen patient care. One of these strategies is the use of gold nanoparticles (GNP) for their multiple properties and their ability to be used as both diagnosis and therapy tools. This review exhaustively details research developing GNPs for use in ophthalmology. The toxicity of GNPs and their distribution in the eye are described through in vitro and in vivo studies. All publications addressing the pharmacokinetics of GNPs administered in the eye are extensively reviewed. In addition, their use as biosensors or for imaging with optical coherence tomography is illustrated. The future of GNPs for ophthalmic therapy is also discussed. GNPs can be used to deliver genes or drugs through different administration routes. Their antiangiogenic and anti-inflammatory properties are of great interest for different ocular pathologies. Finally, GNPs can be used to improve stereotactic radiosurgery, brachytherapy, and photothermal therapy because of their many properties.

Dual Role of the C-Terminal Domain in Osmosensing by Bacterial Osmolyte Transporter ProP.

ProP is a member of the major facilitator superfamily, a proton-osmolyte symporter, and an osmosensing transporter. ProP proteins share extended cytoplasmic carboxyl terminal domains (CTDs) implicated in osmosensing. The CTDs of the best characterized, group A ProP orthologs, terminate in sequences that form intermolecular, antiparallel α-helical coiled coils (e.g., ProPEc, from Escherichia coli). Group B orthologs lack that feature (e.g., ProPXc, from Xanthomonas campestris). ProPXc was expressed and characterized in E. coli to further elucidate the role of the coiled coil in osmosensing. The activity of ProPXc was a sigmoid function of the osmolality in cells and proteoliposomes. ProPEc and ProPXc attained similar activities at the same expression level in E. coli. ProPEc transports proline and glycine betaine with comparable high affinities at low osmolality. In contrast, proline weakly inhibited high-affinity glycine-betaine uptake via ProPXc. The KM for proline uptake via ProPEc increases dramatically with the osmolality. The KM for glycine-betaine uptake via ProPXc did not. Thus, ProPXc is an osmosensing transporter, and the C-terminal coiled coil is not essential for osmosensing. The role of CTD-membrane interaction in osmosensing was examined further. As for ProPEc, the ProPXc CTD co-sedimented with liposomes comprising E. coli phospholipid. Molecular dynamics simulations illustrated association of the monomeric ProPEc CTD with the membrane surface. Comparison with the available NMR structure for the homodimeric coiled coil formed by the ProPEc-CTD suggested that membrane association and homodimeric coiled-coil formation by that peptide are mutually exclusive. The membrane fluidity in liposomes comprising E. coli phospholipid decreased with increasing osmolality in the range relevant for ProP activation. These data support the proposal that ProP activates as cellular dehydration increases cytoplasmic cation concentration, releasing the CTD from the membrane surface. For group A orthologs, this also favors α-helical coiled-coil formation that stabilizes the transporter in an active form.

Deep convolutional networks for quality assessment of protein folds.

Motivation: The computational prediction of a protein structure from its sequence generally relies on a method to assess the quality of protein models. Most assessment methods rank candidate models using heavily engineered structural features, defined as complex functions of the atomic coordinates. However, very few methods have attempted to learn these features directly from the data. Results: We show that deep convolutional networks can be used to predict the ranking of model structures solely on the basis of their raw three-dimensional atomic densities, without any feature tuning. We develop a deep neural network that performs on par with state-of-the-art algorithms from the literature. The network is trained on decoys from the CASP7 to CASP10 datasets and its performance is tested on the CASP11 dataset. Additional testing on decoys from the CASP12, CAMEO and 3DRobot datasets confirms that the network performs consistently well across a variety of protein structures. While the network learns to assess structural decoys globally and does not rely on any predefined features, it can be analyzed to show that it implicitly identifies regions that deviate from the native structure. Availability and implementation: The code and the datasets are available at https://github.com/lamoureux-lab/3DCNN_MQA. Supplementary information: Supplementary data are available at Bioinformatics online.

Cation-π Interactions between Quaternary Ammonium Ions and Amino Acid Aromatic Groups in Aqueous Solution.

Cation-π interactions play important roles in the stabilization of protein structures and protein-ligand complexes. They contribute to the binding of quaternary ammonium ligands (mainly RNH3+ and RN(CH3)3+) to various protein receptors and are likely involved in the blockage of potassium channels by tetramethylammonium (TMA+) and tetraethylammonium (TEA+). Polarizable molecular models are calibrated for NH4+, TMA+, and TEA+ interacting with benzene, toluene, 4-methylphenol, and 3-methylindole (representing aromatic amino acid side chains) based on the ab initio MP2(full)/6-311++G(d,p) properties of the complexes. Whereas the gas-phase affinity of the ions with a given aromatic follows the trend NH4+ > TMA+ > TEA+, molecular dynamics simulations using the polarizable models show a reverse trend in water, likely due to a contribution from the hydrophobic effect. This reversed trend follows the solubility of aromatic hydrocarbons in quaternary ammonium salt solutions, which suggests a role for cation-π interactions in the salting-in of aromatic compounds in solution. Simulations in water show that the complexes possess binding free energies ranging from -1.3 to -3.3 kcal/mol (compared to gas-phase binding energies between -8.5 and -25.0 kcal/mol). Interestingly, whereas the most stable complexes involve TEA+ (the largest ion), the most stable solvent-separated complexes involve TMA+ (the intermediate-size ion).

Mechanism of the Nitric Oxide Dioxygenase Reaction of Mycobacterium tuberculosis Hemoglobin N.

Many globins convert •NO to innocuous NO3- through their nitric oxide dioxygenase (NOD) activity. Mycobacterium tuberculosis fights the oxidative and nitrosative stress imposed by its host (the toxic effects of O2•- and •NO species and their OONO- and •NO2 derivatives) through the action of truncated hemoglobin N (trHbN), which catalyzes the NOD reaction with one of the highest rates among globins. The general NOD mechanism comprises the following steps: binding of O2 to the heme, diffusion of •NO into the heme pocket and formation of peroxynitrite (OONO-), isomerization of OONO-, and release of NO3-. Using quantum mechanics/molecular mechanics free-energy calculations, we show that the NOD reaction in trHbN follows a mechanism in which heme-bound OONO- undergoes homolytic cleavage to give FeIV═O2- and the •NO2 radical but that these potentially harmful intermediates are short-lived and caged by the heme pocket residues. In particular, the simulations show that Tyr33(B10) side chain is shielded from FeIV═O2- and •NO2 (and protected from irreversible oxidation and nitration) by forming stable hydrogen bonds with Gln58(E11) side chain and Leu54(E7) backbone. Aromatic residues Phe46(CD1), Phe32(B9), and Tyr33(B10) promote NO3- dissociation via C-H···O bonding and provide stabilizing interactions for the anion along its egress route.

Loss of hepatic DEPTOR alters the metabolic transition to fasting.

OBJECTIVE: The mechanistic target of rapamycin (mTOR) is a serine/threonine kinase that functions into distinct protein complexes (mTORC1 and mTORC2) that regulates growth and metabolism. DEP-domain containing mTOR-interacting protein (DEPTOR) is part of these complexes and is known to reduce their activity. Whether DEPTOR loss affects metabolism and organismal growth in vivo has never been tested. METHODS: We have generated a conditional transgenic mouse allowing the tissue-specific deletion of DEPTOR. This model was crossed with CMV-cre mice or Albumin-cre mice to generate either whole-body or liver-specific DEPTOR knockout (KO) mice. RESULTS: Whole-body DEPTOR KO mice are viable, fertile, normal in size, and do not display any gross physical and metabolic abnormalities. To circumvent possible compensatory mechanisms linked to the early and systemic loss of DEPTOR, we have deleted DEPTOR specifically in the liver, a tissue in which DEPTOR protein is expressed and affected in response to mTOR activation. Liver-specific DEPTOR null mice showed a reduction in circulating glucose upon fasting versus control mice. This effect was not associated with change in hepatic gluconeogenesis potential but was linked to a sustained reduction in circulating glucose during insulin tolerance tests. In addition to the reduction in glycemia, liver-specific DEPTOR KO mice had reduced hepatic glycogen content when fasted. We showed that loss of DEPTOR cell-autonomously increased oxidative metabolism in hepatocytes, an effect associated with increased cytochrome c expression but independent of changes in mitochondrial content or in the expression of genes controlling oxidative metabolism. We found that liver-specific DEPTOR KO mice showed sustained mTORC1 activation upon fasting, and that acute treatment with rapamycin was sufficient to normalize glycemia in these mice. CONCLUSION: We propose a model in which hepatic DEPTOR accelerates the inhibition of mTORC1 during the transition to fasting to adjust metabolism to the nutritional status.

mTORC1 is Required for Brown Adipose Tissue Recruitment and Metabolic Adaptation to Cold.

In response to cold, brown adipose tissue (BAT) increases its metabolic rate and expands its mass to produce heat required for survival, a process known as BAT recruitment. The mechanistic target of rapamycin complex 1 (mTORC1) controls metabolism, cell growth and proliferation, but its role in regulating BAT recruitment in response to chronic cold stimulation is unknown. Here, we show that cold activates mTORC1 in BAT, an effect that depends on the sympathetic nervous system. Adipocyte-specific mTORC1 loss in mice completely blocks cold-induced BAT expansion and severely impairs mitochondrial biogenesis. Accordingly, mTORC1 loss reduces oxygen consumption and causes a severe defect in BAT oxidative metabolism upon cold exposure. Using in vivo metabolic imaging, metabolomics and transcriptomics, we show that mTORC1 deletion impairs glucose and lipid oxidation, an effect linked to a defect in tricarboxylic acid (TCA) cycle activity. These analyses also reveal a severe defect in nucleotide synthesis in the absence of mTORC1. Overall, these findings demonstrate an essential role for mTORC1 in the regulation of BAT recruitment and metabolism in response to cold.

Effect of Saturated Very Long-Chain Fatty Acids on the Organization of Lipid Membranes: A Study Combining (2)H NMR Spectroscopy and Molecular Dynamics Simulations.

Little is known about the interaction of very long-chain saturated fatty acids (VLCFAs) with biological membranes. However, this could play an important role on interleaflet interactions and signal transduction mechanisms in cells. The aim of this work is to determine how VLCFA structurally adapts in fluid phospholipid bilayers, since both species must exhibit a significant hydrophobic mismatch. The membrane organization has been described by means of (2)H NMR and molecular dynamics simulations. Our results show that the protonation state affects the position and order of free fatty acids (FFAs) in phospholipid membranes. It was shown that the protonated FFA-C24 carboxyl group is located slightly under the POPC head group and therefore its acyl chain can interact with the lipids of the opposite leaflet. This interdigitation of the end of the acyl chain causes a second plateau observed in SC-D profiles, a very unusual feature in lipid systems.

Three-dimensional structure model and predicted ATP interaction rewiring of a deviant RNA ligase 2.

BACKGROUND: RNA ligases 2 are scarce and scattered across the tree of life. Two members of this family are well studied: the mitochondrial RNA editing ligase from the parasitic trypanosomes (Kinetoplastea), a promising drug target, and bacteriophage T4 RNA ligase 2, a workhorse in molecular biology. Here we report the identification of a divergent RNA ligase 2 (DpRNL) from Diplonema papillatum (Diplonemea), a member of the kinetoplastids' sister group. METHODS: We identified DpRNL with methods based on sensitive hidden Markov Model. Then, using homology modeling and molecular dynamics simulations, we established a three dimensional structure model of DpRNL complexed with ATP and Mg2+. RESULTS: The 3D model of Diplonema was compared with available crystal structures from Trypanosoma brucei, bacteriophage T4, and two archaeans. Interaction of DpRNL with ATP is predicted to involve double π-stacking, which has not been reported before in RNA ligases. This particular contact would shift the orientation of ATP and have considerable consequences on the interaction network of amino acids in the catalytic pocket. We postulate that certain canonical amino acids assume different functional roles in DpRNL compared to structurally homologous residues in other RNA ligases 2, a reassignment indicative of constructive neutral evolution. Finally, both structure comparison and phylogenetic analysis show that DpRNL is not specifically related to RNA ligases from trypanosomes, suggesting a unique adaptation of the latter for RNA editing, after the split of diplonemids and kinetoplastids. CONCLUSION: Homology modeling and molecular dynamics simulations strongly suggest that DpRNL is an RNA ligase 2. The predicted innovative reshaping of DpRNL's catalytic pocket is worthwhile to be tested experimentally.

Mechanism of NH4(+) Recruitment and NH3 Transport in Rh Proteins.

In human cells, membrane proteins of the rhesus (Rh) family excrete ammonium and play a role in pH regulation. Based on high-resolution structures, Rh proteins are generally understood to act as NH3 channels. Given that cell membranes are permeable to gases like NH3, the role of such proteins remains a paradox. Using molecular and quantum mechanical calculations, we show that a crystallographically identified site in the RhCG pore actually recruits NH4(+), which is found in higher concentration and binds with higher affinity than NH3, increasing the efficiency of the transport mechanism. A proton is transferred from NH4(+) to a signature histidine (the only moiety thermodynamically likely to accept a proton) followed by the diffusion of NH3 down the pore. The excess proton is circulated back to the extracellular vestibule through a hydrogen bond network, which involves a highly conserved and functionally important aspartic acid, resulting in the net transport of NH3.

Molecular model of hemoglobin N from Mycobacterium tuberculosis bound to lipid bilayers: a combined spectroscopic and computational study.

A singular aspect of the 2-on-2 hemoglobin structures of groups I and II is the presence of tunnels linking the protein surface to the distal heme pocket, supporting the storage and the diffusion of small apolar ligands to/from the buried active site. As the solubility of apolar ligands is greater in biological membranes than in solution, the association of these proteins with biological membranes may improve the efficiency of ligand capture. As very little is known on this subject, we have investigated the interactions between hemoglobin N (HbN), a group I 2-on-2 hemoglobin from the pathogenic Mycobacterium tuberculosis (Mtb), and biological membranes using both experimental techniques and MD simulations. HbN has a potent nitric oxide dioxygenase activity (HbN-Fe²âº-O2 + •NO + H2O → HbN-Fe³âº-OH2 + NO3⁻) that is thought to protect the aerobic respiration of Mtb from inhibition by •NO. Three different membrane compositions were chosen for the studies, representative of the mycobacterial plasma membrane and the mammalian cell membranes. Both the experimental and the modeling results agreed with each other and allow for a detailed molecular description of HbN in association with membranes of different compositions. The results indicated that HbN is a peripheral protein, and the association with the membranes occurred via the pre-A, G, and H helices. In addition, HbN would be allowed to modulate the binding to the membranes via electrostatic interactions between the lipid membranes and the Asp100 residue. In its membrane-bound form the short tunnel of HbN is oriented toward the membrane interior and the other tunnels point toward the solvent. Such protein orientation would facilitate the uptake of nonpolar substrates from the membrane and the release of products to the solvent. It is interesting to note that the pre-A, G, and H helices are conserved among HbN from a few other Mycobacteria.

Identification of a cholesterol-binding pocket in inward rectifier K(+) (Kir) channels.

Cholesterol is the major sterol component of all mammalian plasma membranes. Recent studies have shown that cholesterol inhibits both bacterial (KirBac1.1 and KirBac3.1) and eukaryotic (Kir2.1) inward rectifier K(+) (Kir) channels. Lipid-sterol interactions are not enantioselective, and the enantiomer of cholesterol (ent-cholesterol) does not inhibit Kir channel activity, suggesting that inhibition results from direct enantiospecific binding to the channel, and not indirect effects of changes to the bilayer. Furthermore, conservation of the effect of cholesterol among prokaryotic and eukaryotic Kir channels suggests an evolutionary conserved cholesterol-binding pocket, which we aimed to identify. Computational experiments were performed by docking cholesterol to the atomic structures of Kir2.2 (PDB: 3SPI) and KirBac1.1 (PDB: 2WLL) using Autodock 4.2. Poses were assessed to ensure biologically relevant orientation and then clustered according to location and orientation. The stability of cholesterol in each of these poses was then confirmed by molecular dynamics simulations. Finally, mutation of key residues (S95H and I171L) in this putative binding pocket found within the transmembrane domain of Kir2.1 channels were shown to lead to a loss of inhibition by cholesterol. Together, these data provide support for this location as a biologically relevant pocket.

Development of Semiempirical Models for Proton Transfer Reactions in Water.

This letter presents a method for the parametrization of semiempirical models for proton transfer reactions in water clusters. Two new models are developed: AM1-W, which is a reparameterization of the classic AM1 model, and AM1PG-W, which is a modified AM1-like model including a pairwise correction to the core repulsion function. Both models show good performance on hydrogen-bonding energies and on proton transfer energy profiles, which are of great importance for proton transfer reactions in large water clusters and in proteins. The parametrization method introduced is general and can be used to develop any other system-specific semiempirical models.

Simulation of Liquid and Supercritical Hydrogen Sulfide and of Alkali Ions in the Pure and Aqueous Liquid.

A polarizable model for hydrogen sulfide (H2S) is optimized based on the experimental properties of the monomer and of the bulk liquid. The model is characterized by rigid SH bonds but flexible HSH angle and the polarizability is based on the Drude oscillator model. Bonded parameters and atomic charges are based on the experimental properties of the gaseous monomer. Atomic Lennard-Jones (LJ) parameters are adjusted based on the density of H2S around the critical point (in the temperature range 363-393 K and pressure range 8.023-10.013 MPa). The model gives binding energies for H2S dimers, trimers, and tetramers in good agreement with ab initio MP2(full)/6-311++G(d,p) results. It shows a liquid structure in very good agreement with neutron diffraction data. The model also gives density, self-diffusion coefficient, heat of vaporization, and dielectric constant of liquid hydrogen sulfide at the normal boiling point in good agreement with experimental data. In addition, the model is transferable to high temperature and pressure conditions, as evidenced from simulations up to 542.2 K and 40 MPa. The model is used in combination with the SWM4-NDP water model, with LJ parameters between the S and O atoms adjusted to reproduce the experimental hydration free energy of H2S. Simulations suggest that, in its first solvation shell, a single H2O molecule is solvated by 10 H2S molecules while a single H2S molecule is solvated by 20.5 H2O molecules. Pair-specific LJ parameters between alkali ions (Li(+), Na(+), K(+), Rb(+), Cs(+)) and the S atom are adjusted to reproduce ab initio binding energies of the ion-H2S pairs at the CCSD(T) level. Simulations based on these parameters show that alkali ions have higher coordination numbers and lower solvation free energies in liquid H2S than in liquid water or liquid ammonia. The model is also used to investigate the preferential solvation of the ions in aqueous solutions with a 10% H2S mole fraction. Results show that the ions are preferentially solvated by water in their first solvation shell but have no significant selectivity to either ligands in their second shells.