Publications

In this paper, we propose a new strategy for classifying evaluations in large text corpora, using supervised machine learning (SML). Departing from a conceptual and methodological critique of the use of sentiment measures to recognize object-specific evaluations, we argue that a key challenge consists in determining whether a semantic relationship exists between evaluative expressions and evaluated objects. Regarding sentiment terms as merely potentially evaluative expressions, we thus use a SML classifier to decide whether recognized terms have an evaluative function in relation to the evaluated object. We train and test our classifier on a corpus of 10,004 segments of election coverage from 16 major U.S. news outlets and Tweets by 10 prominent U.S. politicians and journalists. Specifically, we focus on evaluations of political predictions about the outcomes and implications of the 2016 and 2020 U.S. presidential elections. We show that our classifier consistently outperforms both off-the-shelf sentiment tools and a pre-trained transformer-based sentiment classifier. Critically, our classifier correctly discards numerous non-evaluative uses of common sentiment terms, whose inclusion in conventional analyses generates large amounts of false positives. We discuss contributions of our approach to the measurement of object-specific evaluations and highlight challenges for future research.

How does living on property taken from others affect voting behavior? Recent studies argued that benefiting from historical violence leads to support for the far right. We extend this literature with new theoretical insights and data from Israel, using case-specific variation in the nature of displacement to uncover heterogeneous treatment effects. Exploiting the coercion during the settlement of Jewish migrants on rural lands following the 1948 war, we show that living on lands taken from Palestinians consistently led to hawkish right-wing voting in the following 70 years. We also show that exposure to the ruins of the displaced villages increased right-wing voting and that the impact of intergroup contact is divergent: it decreased intolerant voting in most villages but increased it among Jewish communities that reside on violently taken land. Our results are robust when matching is used to account for several controls and spatiotemporal dependencies.

Lipid nanodiscs are nanometric bilayer patches enveloped by confining structures, commonly composed of membrane scaffolding proteins (MSPs). To resolve the interplay between MSP geometry, lipid confinement, and membrane material properties on the nanodisc shape, we apply a continuum elastic theory accounting for lipid bending, tilting, and area deformations. The equilibrium nanodisc shape is then determined by minimizing the elastic free energy functional. Analytic expressions derived under simplifying assumptions demonstrate that the nanodisc shape is sensitive to its size, lipid density, and the lipid tilt and thickness imposed at the contact with the MSP. Under matching physical parameters, these expressions quantitatively reproduce the shape of nanodiscs seen in molecular dynamics simulations, but only if lipid tilt is explicitly considered. We further demonstrate how the bending rigidity can be extracted from the membrane shape profile by fitting the numerically minimized full elastic functional to the membrane shape found in simulations. This fitting procedure faithfully informs on the bending rigidity of nanodiscs larger than ca. 5 nm in radius. The fitted profiles accurately reproduce the increase in bending modulus found using real-space fluctuation analysis of simulated nanodiscs and, for large nanodiscs, also accurately resolve its spatial variations. Our study shows how deformations in lipid patches confined in nanodiscs can be well described by a continuum elastic theory and how this fit can be used to determine local material properties from shape analysis of nanodiscs in simulations. This methodology could potentially allow direct determination of lipid properties from experiments, for example cryo-electron microscopy images of lipid nanodiscs, thereby allowing to guide the development of future nanodisc formulations with desired properties.

Correlations between entangled photons are a key ingredient for testing fundamental aspects of quantum mechanics and an invaluable resource for quantum technologies. However, scattering from a dynamic medium typically scrambles and averages out such correlations. Here we show that multiply-scattered entangled photons reflected from a dynamic complex medium remain partially correlated. We observe in experiments and in full-wave simulations enhanced correlations, within an angular range determined by the transport mean free path, which prevail disorder averaging. Theoretical analysis reveals that this enhancement arises from the interference between scattering trajectories, in which the photons leave the sample and are then virtually reinjected back into it. These paths are the quantum counterpart of the paths that lead to the coherent backscattering of classical light. This work points to opportunities for entanglement transport despite dynamic multiple scattering in complex systems.

Deep eutectic solvents (DESs) show promise in pharmaceutical applications, most prominently as excellent solubilizers. Yet, because DES are complex multi-component mixtures, it is challenging to dissect the contribution of each component to solvation. Moreover, deviations from the eutectic concentration lead to phase separation of the DES, making it impractical to vary the ratios of components to potentially improve solvation. Water addition alleviates this limitation as it significantly decreases the melting temperature and stabilizes the DES single-phase region. Here, we follow the solubility of β-cyclodextrin (β-CD) in DES formed by the eutectic 2:1 mole ratio of urea and choline chloride (CC). Upon water addition to DES, we find that at almost all hydration levels, the highest β-CD solubility is achieved at DES compositions that are shifted from the 2:1 ratio. At higher urea to CC ratios, due to the limited solubility of urea, the optimum composition allowing the highest β-CD solubility is reached at the DES solubility limit. For mixtures with higher CC concentration, the composition allowing optimal solvation varies with hydration. For example, β-CD solubility at 40 wt% water is enhanced by a factor of 1.5 for a 1:2 urea to CC mole ratio compared with the 2:1 eutectic ratio. We further develop a methodology allowing us to link the preferential accumulation of urea and CC in the vicinity of β-CD to its increased solubility. The methodology we present here allows a dissection of solute interactions with DES components that is crucial for rationally developing improved drug and excipient formulations.

Photoisomerization of azobenzenes from their stable E isomer to the metastable Z state is the basis of numerous applications of these molecules. However, this reaction typically requires ultraviolet light, which limits applicability. In this study, we introduce disequilibration by sensitization under confinement (DESC), a supramolecular approach to induce the E-to-Z isomerization by using light of a desired color, including red. DESC relies on a combination of a macrocyclic host and a photosensitizer, which act together to selectively bind and sensitize E-azobenzenes for isomerization. The Z isomer lacks strong affinity for and is expelled from the host, which can then convert additional E-azobenzenes to the Z state. In this way, the host?photosensitizer complex converts photon energy into chemical energy in the form of out-of-equilibrium photostationary states, including ones that cannot be accessed through direct photoexcitation. Chemists often strive to push reactions metaphorically uphill toward less energetically favorable products. The challenge is to keep those products from rolling right back down. Gemen et al. report a clever tactic for twisting azobenzene into its higher-energy Z conformation. Specifically, they lured the more stable E isomer into a supramolecular host, along with a photosensitizer. When visible light injects energy to induce the twist, the Z isomer no longer fits in the cavity, so it gets pushed out before more light can twist it back. ?Jake S. Yeston Spatial constraints in a supramolecular host selectively convert azobenzenes to their metastable state under visible light.Photoisomerization of azobenzenes from their stable E isomer to the metastable Z state is the basis of numerous applications of these molecules. However, this reaction typically requires ultraviolet light, which limits applicability. In this study, we introduce disequilibration by sensitization under confinement (DESC), a supramolecular approach to induce the E-to-Z isomerization by using light of a desired color, including red. DESC relies on a combination of a macrocyclic host and a photosensitizer, which act together to selectively bind and sensitize E-azobenzenes for isomerization. The Z isomer lacks strong affinity for and is expelled from the host, which can then convert additional E-azobenzenes to the Z state. In this way, the host?photosensitizer complex converts photon energy into chemical energy in the form of out-of-equilibrium photostationary states, including ones that cannot be accessed through direct photoexcitation. Chemists often strive to push reactions metaphorically uphill toward less energetically favorable products. The challenge is to keep those products from rolling right back down. Gemen et al. report a clever tactic for twisting azobenzene into its higher-energy Z conformation. Specifically, they lured the more stable E isomer into a supramolecular host, along with a photosensitizer. When visible light injects energy to induce the twist, the Z isomer no longer fits in the cavity, so it gets pushed out before more light can twist it back. ?Jake S. Yeston Spatial constraints in a supramolecular host selectively convert azobenzenes to their metastable state under visible light.

Virus-like particles (VLPs) are noninfectious nanocapsules that can be used for drug delivery or vaccine applications. VLPs can be assembled from virus capsid proteins around a condensing agent, such as RNA, DNA, or a charged polymer. Electrostatic interactions play an important role in the assembly reaction. VLPs assemble from many copies of capsid protein, with a combinatorial number of intermediates. Hence, the mechanism of the reaction is poorly understood. In this paper, we combined solution small-angle X-ray scattering (SAXS), cryo-transmission electron microscopy (TEM), and computational modeling to determine the effect of ionic strength on the assembly of Simian Vacuolating Virus 40 (SV40)-like particles. We mixed poly(styrene sulfonate) with SV40 capsid protein pentamers at different ionic strengths. We then characterized the assembly product by SAXS and cryo-TEM. To analyze the data, we performed Langevin dynamics simulations using a coarse-grained model that revealed incomplete, asymmetric VLP structures consistent with the experimental data. We found that close to physiological ionic strength, $$T=1$$VLPs coexisted with VP1 pentamers. At lower or higher ionic strengths, incomplete particles coexisted with pentamers and $$T=1$$particles. Including the simulated structures was essential to explain the SAXS data in a manner that is consistent with the cryo-TEM images.

We investigated how the self-association of isolated tubulin dimers affects the rate of GTP hydrolysis and the equilibrium of nucleotide exchange. Both reactions are relevant for microtubule (MT) dynamics. We used HPLC to determine the concentrations of GDP and GTP and thereby the GTPase activity of SEC-eluted tubulin dimers in assembly buffer solution, free of glycerol and tubulin aggregates. When GTP hydrolysis was negligible, the nucleotide exchange mechanism was studied by determining the concentrations of tubulin-free and tubulin-bound GTP and GDP. We observed no GTP hydrolysis below the critical conditions for MT assembly (either below the critical tubulin concentration and/or at low temperature), despite the assembly of tubulin 1D curved oligomers and single-rings, showing that their assembly did not involve GTP hydrolysis. Under conditions enabling spontaneous slow MT assembly, a slow pseudo-first-order GTP hydrolysis kinetics was detected, limited by the rate of MT assembly. Cryo-TEM images showed that GTP-tubulin 1D oligomers were curved also at 36 °C. Nucleotide exchange depended on the total tubulin concentration and the molar ratio between tubulin-free GDP and GTP. We used a thermodynamic model of isodesmic tubulin self-association, terminated by the formation of tubulin single-rings to determine the molar fractions of dimers with exposed and buried nucleotide exchangeable sites (E-sites). Our analysis shows that the GDP to GTP exchange reaction equilibrium constant was an order-of-magnitude larger for tubulin dimers with exposed E-sites than for assembled dimers with buried E-sites. This conclusion may have implications on the dynamics at the tip of the MT plus end.