Artificial intelligence in virtual screening: Models versus experiments.

A typical drug discovery project involves identifying active compounds with significant binding potential for selected disease-specific targets. Experimental high-throughput screening (HTS) is a traditional approach to drug discovery, but is expensive and time-consuming when dealing with huge chemical libraries with billions of compounds. The search space can be narrowed down with the use of reliable computational screening approaches. In this review, we focus on various machine-learning (ML) and deep-learning (DL)-based scoring functions developed for solving classification and ranking problems in drug discovery. We highlight studies in which ML and DL models were successfully deployed to identify lead compounds for which the experimental validations are available from bioassay studies.

Structure-based drug repurposing: Traditional and advanced AI/ML-aided methods.

The current global health emergency in the form of the Coronavirus 2019 (COVID-19) pandemic has highlighted the need for fast, accurate, and efficient drug discovery pipelines. Traditional drug discovery projects relying on in vitro high-throughput screening (HTS) involve large investments and sophisticated experimental set-ups, affordable only to big biopharmaceutical companies. In this scenario, application of efficient state-of-the-art computational methods and modern artificial intelligence (AI)-based algorithms for rapid screening of repurposable chemical space [approved drugs and natural products (NPs) with proven pharmacokinetic profiles] to identify the initial leads is a powerful option to save resources and time. Structure-based drug repurposing is a popular in silico repurposing approach. In this review, we discuss traditional and modern AI-based computational methods and tools applied at various stages for structure-based drug discovery (SBDD) pipelines. Additionally, we highlight the role of generative models in generating molecules with scaffolds from repurposable chemical space.

Molecular basis for stereoselective transport of fenoterol by the organic cation transporters 1 and 2.

Stereoselectivity is important in many pharmacological processes but its impact on drug membrane transport is scarcely understood. Recent studies showed strong stereoselective effects in the cellular uptake of fenoterol by the organic cation transporters OCT1 and OCT2. To provide possible molecular explanations, homology models were developed and the putative interactions between fenoterol enantiomers and key residues explored in silico through computational docking, molecular dynamics simulations, and binding free energy calculations as well as in vitro by site-directed mutagenesis and cellular uptake assays. Our results suggest that the observed 1.9-fold higher maximum transport velocity (vmax) for (R,R)- over (S,S)-fenoterol in OCT1 is because the enantiomers bind to two distinct binding sites. Mutating PHE355 and ILE442, predicted to interact with (R,R)-fenoterol, reduced the vmax ratio to 1.5 and 1.3, respectively, and to 1.2 in combination. Mutating THR272, predicted to interact with (S,S)-fenoterol, slightly increased stereoselectivity (vmax ratio of 2.2), while F244A resulted in a 35-fold increase in vmax and a lower affinity (29-fold higher Km) for (S,S)-fenoterol. Both enantiomers of salbutamol, for which almost no stereoselectivity was observed, were predicted to occupy the same binding pocket as (R,R)-fenoterol. Unlike for OCT1, both fenoterol enantiomers bind in the same region in OCT2 but in different conformations. Mutating THR246, predicted to interact with (S,S)-fenoterol in OCT2, led to an 11-fold decreased vmax. Altogether, our mutagenesis results correlate relatively well with our computational predictions and thereby provide an experimentally-corroborated hypothesis for the strong and contrasting enantiopreference in fenoterol uptake by OCT1 and OCT2.

Unusual binding-site-specific photophysical properties of a benzothiazole-based optical probe in amyloid beta fibrils.

Optical imaging of amyloid fibrils serves as a cost-effective route for the diagnosis of Alzheimer-like conformational diseases. However, the challenge here is to optimize the binding affinity and photophysical properties of the optical imaging agents in a way specific to certain types of amyloids. In a few occasions it is shown that novel optical imaging agents can be designed to bind to a particular type of amyloid fibril with larger binding affinity and specificity. There is also a recent report on photoluminescent polythiophenes which display photophysical properties that can be used to distinguish the variants or subtypes of amyloids (J. Rasmussen et al., Proc. Natl. Acad. Sci. U. S. A., 2017, 114(49), 13018-13023). Based on a multiscale modeling approach, here, we report on the complementary aspect that the photophysical properties of a benzothiazole based optical probe (referred to as BTA-3) can be specific to the binding sites in the same amyloid fibrils and we attribute this to its varying electronic structure in different sites. As reported experimentally from competitive binding assay studies for many amyloid staining molecules and tracers, we also show multiple binding sites in amyloid fibrils for this probe. In particular, BTA-3 displayed a red-shift in its low-frequency absorption band only in site-4, a surface site of amyloid fibrils when compared to the spectra in water solvent. In the remaining sites, it exhibited a less significant blue shift for the same absorption band.

Elucidating the Mechanism of Zn(2+) Sensing by a Bipyridine Probe Based on Two-Photon Absorption.

In this work, we examine, by means of computational methods, the mechanism of Zn(2+) sensing by a bipyridine-centered, D-π-A-π-D-type ratiometric molecular probe. According to recently published experimental data [Divya, K. P.; Sreejith, S.; Ashokkumar, P.; Yuzhan, K.; Peng, Q.; Maji, S. K.; Tong, Y.; Yu, H.; Zhao, Y.; Ramamurthy, P.; Ajayaghosh, A. A ratiometric fluorescent molecular probe with enhanced two-photon response upon Zn(2+) binding for in vitro and in vivo bioimaging. Chem. Sci. 2014, 5, 3469-3474], after coordination to zinc ions the probe exhibits a large enhancement of the two-photon absorption cross section. The goal of our investigation was to elucidate the mechanism behind this phenomenon. For this purpose, linear and nonlinear optical properties of the unbound (cation-free) and bound probe were calculated, including the influence of solute-solvent interactions, implicitly using a polarizable continuum model and explicitely employing the QM/MM approach. Because the results of the calculations indicate that many conformers of the probe are energetically accessible at room temperature in solution and hence contribute to the signal, structure-property relationships were also taken into account. Results of our simulations demonstrate that the one-photon absorption bands for both the unbound and bound forms correspond to the bright π → π* transition to the first excited state, which, on the other hand, exhibits negligible two-photon activity. On the basis of the results of the quadratic response calculations, we put forward a notion that it is the second excited state that gives the strong signal in the experimental nonlinear spectrum. To explain the differences in the two-photon absorption activity for the two lowest-lying excited states and nonlinear response enhancement upon binding, we employed the generalized few-state model including the ground, first, and second excited states. The analysis of the optical channel suggests that the large two-photon response is due to the coordination-induced increase of the transition moment from the first to the second excited state.

Studies on an on/off-switchable immunosensor for troponin T.

Regeneration is a key goal in the design of immunosensors. In this study, we report the temperature-regulated interaction of N-isopropylacrylamide (PNIPAAm) functionalised cardiac troponin T (cTnT) with anti-cTnT. Covalently bonded PNIPAAm on an anti-cTnT bioelectrode showed on/off-switchability, regeneration capacity and temperature triggered sensitivity for cTnT. Above the lower critical solution temperature (LCST), PNIPAAm provides a liphophilic microenvironment with specific volume reduction at the bioelectrode surface, making available binding space for cTnT, and facilitating analyte recognition. Computational studies provide details about the structural changes occurring at the electrode above and below the LCST. Furthermore, free energies associated with the binding of cTnT with PNIPAAm at 25 (ΔGcoil=-6.0 Kcal/mole) and 37 °C (ΔGglobular=-41.0 kcal/mole) were calculated to elucidate the interaction and stability of the antigen-antibody complex. The responsiveness of such assemblies opens the way for miniaturised, smart immuno-technologies with 'built-in' programmable interactions of antigen-antibody upon receiving stimuli.

In silico modeling and experimental evidence of coagulant protein interaction with precursors for nanoparticle functionalization.

The design of novel protein-nanoparticle hybrid systems has applications in many fields of science ranging from biomedicine, catalysis, water treatment, etc. The main barrier in devising such tool is lack of adequate information or poor understanding of protein-ligand chemistry. Here, we establish a new strategy based on computational modeling for protein and precursor linkers that can decorate the nanoparticles. Moringa oleifera (MO2.1) seed protein that has coagulation and antimicrobial properties was used. Superparamagnetic nanoparticles (SPION) with precursor ligands were used for the protein-ligand interaction studies. The molecular docking studies reveal that there are two binding sites, one is located at the core binding site; tetraethoxysilane (TEOS) or 3-aminopropyl trimethoxysilane (APTES) binds to this site while the other one is located at the side chain residues where trisodium citrate (TSC) or Si60 binds to this site. The protein-ligand distance profile analysis explains the differences in functional activity of the decorated SPION. Experimentally, TSC-coated nanoparticles showed higher coagulation activity as compared to TEOS- and APTES-coated SPION. To our knowledge, this is the first report on in vitro experimental data, which endorses the computational modeling studies as a powerful tool to design novel precursors for functionalization of nanomaterials; and develop interface hybrid systems for various applications.

Hybrid density functional-molecular mechanics calculations for core-electron binding energies of glycine in water solution.

We report hybrid density functional theory-molecular mechanics (DFT/MM) calculations performed for glycine in water solution at different pH values. In this paper, we discuss several aspects of the quantum mechanics-molecular mechanics (QM/MM) simulations where the dynamics and spectral binding energy shifts are computed sequentially, and where the latter are evaluated over a set of configurations generated by molecular or Car-Parrinello dynamics simulations. In the used model, core ionization takes place in glycine as a quantum mechanical (QM) system modeled with DFT, and the solution is described with expedient force fields in a large molecular mechanical (MM) volume of water molecules. The contribution to the core electronic binding energy from all interactions within and between the two (DFT and MM) parts is accounted for, except charge transfer and dispersion. While the obtained results were found to be in qualitative agreement with experiment, their precision must be qualified with respect to the problem of counter ions, charge transfer and optimal division of QM and MM parts of the system. Results are compared to those of a recent study [Ottoson et al., J. Am. Chem. Soc., 2011, 133, 3120].

Hybrid density functional theory/molecular mechanics calculations of two-photon absorption of dimethylamino nitro stilbene in solution.

The dimethylamino nitro stilbene (DANS) molecule is studied for exploring solvent effects on two-photon absorption using the quantum mechanical/molecular mechanical (QM/MM) response theory approach, where the quantum part is represented by density functional theory. We have explored the role of geometrical change of the chromophore in solution, the importance of taking a dynamical average over the sampled structures and the role of a granular representation of the polarization and electrostatic interactions with the classically described medium. The line shape function was simulated by the QM/MM technique thereby allowing for non-empirical prediction of the absolute two-photon cross section. We report a maximum in the TPA cross section for a medium of intermediate solvent polarity (i.e. in chloroform) and provide the grounds for an explanation of this effect which recently has been experimentally observed for a series of charge transfer species in solvents of different polarity. The calculations of absorption energies reproduce well the positive solvatochromic behavior of DANS and are in good agreement with experimental spectra available for the chloroform and DMSO solvents. In line with recent development of the QM/MM response technique for color modeling, we find this methodology to offer a versatile tool to predict and analyze two-photon absorption phenomena taking place within a medium.

Solvation shell structure of cyclooctylpyranone in water solvent and its comparative structure, dynamics and dipole moment in HIV protease.

We have investigated the solvation structure for cyclooctylpyranone (COP) in water solvent using force-field molecular dynamics (MD) and Car-Parrinello mixed quantum mechanics-molecular mechanics (CPMD) calculations. The MD calculations show that in water solvent COP can exist in two conformational states which differ with respect to the relative orientations of the three rings, namely phenyl, pyranone and cyclooctane. We report the existence of strong orientational preference for the water molecule in the first solvation shell and the orientational preference disappears for solvent molecules beyond the first solvation shell. In order to investigate the confinement effect on the structure, dynamics, charge distribution and dipole moment of COP, we have carried out MD and CPMD calculations for COP within HIV type-1 protease (PR). Interestingly, we do not see any conformational transitions for COP within the protein cavity and it remains as a single conformer. We do see a remarkable effect of confinement on few other torsional degrees of freedom such as gg to tg conformational shift for the propyl group of COP. However, the methyl group rotational dynamics remains similar in the water solvent and in the protein environment. Also, within the protein cavity, the COP molecule is more polarized when compared to water solvent. Static ab initio electronic structure calculations were performed on the COP molecule with varying torsional angle in order to investigate the angle dependence of the molecular volume and energy.

What does pressure decide to cook with orientationally disordered plastic phase of cubane: an orientational glass or crystal?

The effect of pressure on the structure and reorientational motion of molecules in orientationally disordered (OD) crystalline phase of cubane has been investigated in detail using variable shape molecular simulations in constant-pressure constant-temperature ensemble. Complete orientational ordering occurs at a pressure of 1.0 GPa and the OD phase transforms to an orientationally ordered phase at this pressure. The transition is associated with a kink in the variation of structural parameters such as cell parameters, unit-cell volume, and interaction energy. This transition is also associated with an anomaly in specific heat. Above this transition pressure, the structural quantities display only smaller changes with further increase in pressure. The structure of high-pressure orientationally ordered (HPOO) phase has been characterized using radial distribution functions and orientational distribution function. From detailed analysis of the structure of HPOO phase we conclude that it is isostructural with low-temperature orientationally ordered phase. The OD phase has four times larger compressibility than the HPOO phase.