COVID-19 therapy optimization by AI-driven biomechanical simulations.

The COVID-19 disease causes pneumonia in many patients that in the most serious cases evolves into the Acute Distress Respiratory Syndrome (ARDS), requiring assisted ventilation and intensive care. In this context, identification of patients at high risk of developing ARDS is a key point for early clinical management, better clinical outcome and optimization in using the limited resources available in the intensive care units. We propose an AI-based prognostic system that makes predictions of oxygen exchange with arterial blood by using as input lung Computed Tomography (CT), the air flux in lungs obtained from biomechanical simulations and Arterial Blood Gas (ABG) analysis. We developed and investigated the feasibility of this system on a small clinical database of proven COVID-19 cases where the initial CT and various ABG reports were available for each patient. We studied the time evolution of the ABG parameters and found correlation with the morphological information extracted from CT scans and disease outcome. Promising results of a preliminary version of the prognostic algorithm are presented. The ability to predict the evolution of patients' respiratory efficiency would be of crucial importance for disease management.

Curvature dynamics and long-range effects on fluid-fluid interfaces with colloids.

We investigate the dynamics of a phase-separating binary fluid, containing colloidal dumbbells anchored to the fluid-fluid interface. Extensive lattice Boltzmann-immersed boundary method simulations reveal that the presence of soft dumbbells can significantly affect the curvature dynamics of the interface between phase-separating fluids, even though the coarsening dynamics is left nearly unchanged. In addition, our results show that the curvature dynamics exhibits distinct non-local effects, which might be exploited for the design of new soft mesoscale materials. We point out that the inspection of the statistical dynamics of the curvature can disclose new insights into local inhomogeneities of the binary fluid configuration, as a function of the volume fraction and aspect ratio of the dumbbells.

The effect of protein composition on hydration dynamics.

Water dynamics at the surface of two homologous proteins with different thermal resistances is found to be unaffected by the different underlying amino-acid compositions, and when proteins are folded it responds similarly to temperature variations. Upon unfolding the water dynamics slowdown with respect to bulk decreases by a factor of two. Our findings are explained by the dominant topological perturbation induced by the protein on the water hydrogen bond dynamics.

Three-band decomposition analysis of wall shear stress in pulsatile flows.

Space-time patterns of wall shear stress (WSS) resulting from the numerical simulation of pulsating hemodynamic flows in semicoronal domains are analyzed, in the case of both regular semicoronal domains and semicoronal domains with bumpy insertions, mimicking aneurysm-like geometries. A new family of cardiovascular risk indicators, which we name three-band diagrams (TBDs), are introduced, as a sensible generalization of the two standard indicators, i.e., the time-averaged WSS and the oscillatory shear index. TBDs provide a handy access to additional information contained in the dynamic structure of the WSS signal as a function of the physiological risk threshold, thereby allowing a quick visual assessment of the risk sensitivity to individual fluctuations of the physiological risk thresholds. Due to its generality, TBD analysis is expected to prove useful for a wide host of applications in science, engineering, and medicine, where risk assessment plays a central role.

Pressure-induced core packing and interfacial dehydration in nonionic C12E6 micelle in aqueous solution.

A spherical micelle of C12E6 is simulated at different pressures, from 0.001 to 3 kbar, by molecular dynamics. On increasing the pressure the alkyl tails of the surfactants pack tightly and stretch. At 3 kbar we observe dynamical slowing down of the oil core of the micelle. At that pressure the core is characterized by a high oil density, rho oil approximately 0.85 g/cm(3), regular density oscillation, and low chain entropy. Pressure affects the interfacial region as well. Dehydration, induced by the collapse of the hydrophilic head groups, is observed in the inner part of the interface. Such dehydration resembles temperature dehydration but differs in details. Our results support the interpretation of recent experiments on micellar solutions at high pressure.

IMAGE: a new tool for the prediction of transcription factor binding sites.

IMAGE is an application tool, based on the vector quantization method, aiding the discovery of nucleotidic sequences corresponding to Transcription Factor binding sites. Starting from the knowledge of regulation regions of a number of co-expressed genes, the software is able to predict the occurrence of specific motifs of different lengths (starting from 6 base pairs) with a defined number of punctual mutations.

Molecular dynamics simulation of ratchet motion in an asymmetric nanochannel.

The persistence of ratchet effects, i.e., nonzero mass flux under a zero-mean time-dependent drive, when many-body interactions are present, is studied via molecular dynamics (MD) simulations of a simple liquid flowing in an asymmetric nanopore. The results show that (i) ratchet effects persist under many-body density correlations induced by the forcing; (ii) two distinct linear responses (flux proportional to the drive amplitude) appear under strong loads. One regime has the same conductivity of linear response theory up to a forcing of about 10 kT, while the second displays a smaller conductivity, the difference in responses is due to geometric effects alone. (iii) Langevin simulations based on a naive mapping of the many-body equilibrium bulk diffusivity, D, onto the damping rate, gamma are also found to yield two distinct linear responses. However, in both regimes, the flux is significantly smaller than the one of MD simulations.

The role of water coordination in binary mixtures. A study of two model amphiphilic molecules in aqueous solutions by molecular dynamics and NMR.

Two binary aqueous mixtures which contain the small amphiphilic molecules TMAO (trimethylamine-N-oxide) and TBA (tert-butyl alcohol) have been investigated by molecular dynamics simulations and NMR chemical shift and self-diffusion measurements. TMAO is an osmolyte, while TBA is a monohydrate alcohol. Both possess bulky hydrophobic groups and polar heads, namely, NO in TMAO and OH in TBA. The hydrophilic/hydrophobic content of these isosteric molecules strongly modulates the structure and dynamics of the hydration shell, which is thought to be responsible for the effects observed on proteins and phospholipids. Simulation results, especially on hydrogen-bond networking, spatial correlations, and self-diffusivity, are consistent with NMR data and agree well with previous numerical studies on similar solutions. The methods employed allow the elucidation of the microscopic features of the solutions. For TBA solutions, the hydration shell is found to have a low density and a large spatial spread, and thus, above the molar fraction of 0.03, reduction of hydrophobic hydration drives self-aggregation of the solute. This effect does not take place in TMAO solutions, where the hydration shell is more compact and stable, maintaining its structure over a wider range of solute concentrations.

Accurate calculation of three-body depletion interactions.

We compute three-body depletion interactions in a hard-sphere mixture within the framework of density-functional theory and by considering the infinite dilution limit of the functional. The results look very accurate and show three-body interactions much smaller than the pair depletion ones, revealing that these are strongly influenced by correlations and have a decay length similar to the two-body depletion potential. The results are compared with the predictions of the Asakura-Oosawa model for the triplet interactions.

Molecular characterization of a laminin-derived oligopeptide with implications in biomimetic applications.

The molecular properties of the laminin-derived oligopeptide, H-CDPGYIGSR-NH2, have been investigated with the aid of a tandem computer simulation/experimental approach. The simulation studies placed a particular emphasis on studying the oligopeptide in aqueous media, as well as in a grafted or immobilized state. The simulations revealed the presence of a stable double hydrogen bond between arginine (R) and aspartic acid (D) residues. The mutation of the terminal arginine with lysine, another hydrophilic and positively charged amino acid, resulted in a drastic structural change, thus suggesting a major role of the terminal arginine residue in the overall oligopeptide's conformation and, hence, its bioactivity. In addition, the involvement of the aspartic acid residue in overall peptide structural stabilization also illustrates a previously undetermined role for this region (i.e. CDPG) of the oligopeptide. A subsequent in vitro experiment demonstrated a significant loss of bioactivity upon mutating the terminal residue from arginine to lysine, thereby corroborating the overall findings of the computational model.

Constrained systems and statistical distribution

Theoretical advances in the statistical description of non-Hamiltonian systems [Tuckerman, Mundy, and Martyna, Europhys. Lett. 45, 149 (1999)] have recently led us to rethink the definition of the phase-space probability by using the proper invariant measure. Starting from this point of view, we will derive the statistical distribution of constrained systems, considered as non-Hamiltonian, in Cartesian coordinates, in a way independent from the standard Lagrangian treatment. Furthermore, we will analyze the statistical distribution of the Nose-Hoover isothermal (canonical or NVT) dynamics, considering with care the conservation laws hidden in the evolution equations. Consequently, we will correct the equations of motion in order to obtain the proper statistical ensemble. The isobaric-isothermal (constant pressure or NpT) form of the Nose-Hoover dynamics is then considered as a nontrivial extension of the described procedure and, similarly to the NVT case, we will correct the equations of motion with respect to previous versions. The case of a constrained system coupled to a thermostat and a piston is easily handled by means of the same formalism, whereas the Lagrangian treatment becomes involved.

Size selectivity of narrow pores.

The selectivity of micropores and ion channels is examined for simple pore topologies within the framework of density functional theory of highly confined fluids. In an infinite cylindrical pore purely steric (excluded volume) effects are shown to lead to strong, nontrivial size selectivity, which is highly sensitive to the pore radius. A crude modeling of electrostatic effects does not alter the relative absorbance of Na+ and K+ ions in a significant way.

Enhanced sampling of rare events

We present a method to enhance sampling of a given reaction coordinate by projecting part of the random thermal noise along a preferential direction. The approach is promising to study rough energy landscapes and highly activated barriers that can be overcome by increasing the attempt frequency. Furthermore it allows us to rescale a given reaction coordinate without biasing the configurational properties of the system. A major advantage of the method is its simplicity in the analytical derivation and numerical implementation.

Molecular dynamics simulations of Cu,Zn superoxide dismutase: effect of temperature on dimer asymmetry.

Molecular dynamics simulations of solvated dimeric Cu,Zn superoxide dismutase have been carried out at four temperatures, namely 200, 225, 250 and 300 K. Analysis of the backbone-to-backbone hydrogen bonds number indicates that the symmetry observed in the two subunits at 200 K is gradually lost by heating the system. The C(alpha) atoms displacement cross-correlation maps confirm that the asymmetric behaviour of the two subunits increases as a function of temperature. The dynamic cross-correlation of the subunits volumes indicates a fast correlation between the two subunits at 300 K, which is delayed upon lowering the simulation temperature. These results indicate that temperature plays an essential role in injecting such an asymmetry; the two subunits being asymmetric and in rapid communication at 300 K, and almost symmetric and in slow communication at lower temperatures.

Origin of the low-frequency modes of globular proteins.

The incoherent structure factor associated with the nonexchangeable hydrogen atoms of carbomonoxy myoglobin at 80 and 200 K in a solvated sample has been computed from molecular dynamics simulations. The evaluated mean square displacements are almost three times larger than the experimental ones. However, the low-frequency band centered at 3 meV, observed in neutron scattering experiments, is reproduced. Evaluation of the contribution to the spectrum from different protein shells and backbone and side chain atoms allows us to propose that the low-frequency mode around 3 meV, which is a constant feature of all globular proteins, is mainly due to the hydrogen atoms belonging to the inner shell, suggesting that this mode is intrinsic to the protein itself.