Angela Violi

<div class="section abstract"><div class="htmlview paragraph">The complexity and variability of modern aviation fuels necessitate the development of robust and efficient tools to assess their properties accurately, particularly within the certification framework established by the American Society for Testing and Materials (ASTM). Therefore, previous research has developed predictive models to reduce the experimental burden by predicting aviation fuel properties from broad chemical classes. While two-dimensional Gas Chromatography (GC×GC) provides detailed compositional information, it only identifies the weight of hydrocarbon families (aromatics, cycloalkanes, n-alkanes, iso-alkanes), not individual molecules. Aviation fuels are complex, and their composition can contain more than 60 key classes, the majority of which are isomeric. As a result, an exceptionally high number of possible molecule combinations makes random selection prone to high errors in property prediction. To this end, we used a Monte Carlo approach to search for the optimal combination of 64 hydrocarbon molecules from this vast combinatorial space. By exploring up to 500 million combinations, we aim to determine the molecule set that best predicts mass density, kinematic viscosity, and distillation temperature using linear mixing rules. These rules calculate the properties of molecule mixtures using the weight of each molecule in the mixture and the pure molecules’ properties. We used experimental data for various aviation fuels, including conventional jet fuels, sustainable aviation fuels, and rocket propulsion fuels. Results showed that the isomeric effect has a substantial role in predicting mass density, kinematic viscosity, and the distillation temperature. Results showed that the linear mixing rules could outperform machine learning that overlooks the isomeric effect for the three properties. This research benefits the surrogate fuel analysis, which requires defining a surrogate mixture of hydrocarbon molecules, and will provide insights into the best isomers or molecules to choose to predict aviation fuel properties with the least error. This work will help deliver aviation fuel producers with a relatively accurate pre-screening tool for property prediction, minimizing the need for iterative experimental processes.</div></div>

In this work, we introduce a new methodology to combine the available methods to predict the properties of complex hydrocarbon mixtures such as aviation fuels. Due to the complexity of aviation fuels, the available methods perform well individually on some of the experimental observations and vice versa on others when a surrogate aviation fuel is defined and used. To this end, we introduce a new ensemble model based on the existing methods that combine and weigh their predictions. We employ the probabilistic Bayesian approach to predict aviation fuel properties with confidence levels. This is necessary because the available experimental data for aviation fuels is generally limited, which leads to overfitting. We adopt both “interpretable” Bayesian regression and a more “black-box” approach to Bayesian neural networks. An ensemble of predictive methods provided better predictions than the individual methods with robust confidence levels for three properties considered: mass density, kinematic viscosity, and flash point. A significant reduction in the mean absolute percentage error was obtained for mass density predictions, from 1.25% to 0.57% and 0.42%, using the Bayesian linear regression (BLR) and Bayesian Neural Network (BNN), respectively. The error in kinematic viscosity predictions was reduced from 17.25% to 9.02% and 6.79% using BLR and BNN, respectively. The error in flash point predictions is reduced from 9.04% to 5.83% by BLR and to 5.51% by BNN. The importance of the methods in the ensemble did not fully follow their individual performance, where the accurate models may not be the most important. The ensemble approach allows for the inclusion of new methods, even if they are slightly less accurate. This methodology can be extended to predict other aviation fuel properties and incorporate any predictive model. It also offers a way to generate valid training data for generative Artificial Intelligence (AI) models, helping to address the scarcity of aviation fuel data. • The study introduces the concept of using an ensemble of property predictive models to predict aviation fuel properties. • The uncertainty of predictions via the Bayesian approach is crucial since the aviation fuel data is often limited, making the predictive models prone to overfitting. • There is a tradeoff between accuracy and uncertainty, where a balanced perspective is required. • The most accurate predictive models in the ensemble are not necessarily the most important.

While percentages of women employed in STEM fields in the United States have generally risen, albeit slowly, over the past several decades, the percentages of women employed in engineering fields specifically has increased at a glacial and stagnating pace. According to the Bureau of Labor Statistics, only 3% of practicing engineers were women in the 1970s compared to about 16% in 2023. There is slightly more growth in the percentages of women graduating with engineering degrees, with current numbers hovering around 24% for undergraduate and 26% for graduate students across all areas of engineering; however, this growth has plateaued in the past decade. Women make up fewer than 20% of graduates in Mechanical Engineering in both undergraduate and graduate degrees, a field historically associated with heavy industry, which may contribute to this disparity. This paper aims to surface and explore aspects of these trends, laying the groundwork for a larger book project that will share the stories of women in the Mechanical Engineering department at the University of Michigan, Ann Arbor. We will summarize the trends of women in both the engineering workforce and in engineering academia. We will delve into the data for Mechanical Engineering relative to other fields and summarize reasons the percentages of women in Mechanical Engineering programs and occupations have stalled.

We propose in this work a new methodology for predicting aviation fuel properties by combining the predictions of existing methods into an ensemble. The performance of current methods varies among different types of aviation fuels, including conventional jet fuels, Sustainable Aviation Fuels (SAF), and biofuels. Aviation fuels are a complex mixture of several hydrocarbons, which may also lead to high prediction errors. The predictions of the methods in the ensemble are weighted, where the weights are obtained using a probabilistic Bayesian linear regression (BLR) model. We tested the methodology used to predict the mass density of different aviation fuels. We also compared the results of BLR with those of deterministic optimization using Particle Swarm Optimization (PSO). Both BLR and PSO reduced the prediction error of mass density from 1.24% to 0.57% and 0.49%, respectively. We noticed a trade-off between uncertainty and accuracy in BLR results, where higher accuracy is accompanied by higher uncertainty. Accordingly, more accurate results provided by PSO or possibly other deterministic approaches may be misleading, and uncertainty should be considered to draw more robust conclusions. Results also showed that the individual performance of the methods does not follow their importance in the ensemble, i.e., the most accurate models do not necessarily have the highest weights; less accurate models could be more important. As a result, new predicting methods can be added even if they are, to some extent, less accurate than the methods in the ensemble.

Abstract Nanoparticles formed in nonthermal plasmas (NTPs) can have unique properties and applications. However, modeling their growth in these environments presents significant challenges due to the non-equilibrium nature of NTPs, making them computationally expensive to describe. In this work, we address the challenges associated with accelerating the estimation of parameters needed for these models. Specifically, we explore how different machine learning models can be tailored to improve prediction outcomes. We apply these methods to reactive classical molecular dynamics data, which capture the processes associated with colliding silane fragments in NTPs. These reactions exemplify processes where qualitative trends are clear, but their quantification is challenging, hard to generalize, and requires time-consuming simulations. Our results demonstrate that good prediction performance can be achieved when appropriate loss functions are implemented and correct invariances are imposed. While the diversity of molecules used in the training set is critical for accurate prediction, our findings indicate that only a fraction (15%–25%) of the energy and temperature sampling is required to achieve high levels of accuracy. This suggests a substantial reduction in computational effort is possible for similar systems.

Nanoparticles (NPs) formed in nonthermal plasmas (NTPs) can have unique properties and applications. However, modeling their growth in these environments presents significant challenges due to the non-equilibrium nature of NTPs, making them computationally expensive to describe. In this work, we address the challenges associated with accelerating the estimation of parameters needed for these models. Specifically, we explore how different machine learning models can be tailored to improve prediction outcomes. We apply these methods to reactive classical molecular dynamics data, which capture the processes associated with colliding silane fragments in NTPs. These reactions exemplify processes where qualitative trends are clear, but their quantification is challenging, hard to generalize, and requires time-consuming simulations. Our results demonstrate that good prediction performance can be achieved when appropriate loss functions are implemented and correct invariances are imposed. While the diversity of molecules used in the training set is critical for accurate prediction, our findings indicate that only a fraction (15-25\%) of the energy and temperature sampling is required to achieve high levels of accuracy. This suggests a substantial reduction in computational effort is possible for similar systems.

Tree ensembles achieve state-of-the-art performance on numerous prediction tasks. We propose Joint Optimization of Piecewise Linear Ensembles (JOPLEn), which jointly fits piecewise linear models at all leaf nodes of an existing tree ensemble. In addition to enhancing the ensemble expressiveness, JOPLEn allows several common penalties, including sparsity-promoting and subspace-norms, to be applied to nonlinear prediction. For example, JOPLEn with a nuclear norm penalty learns subspace-aligned functions. Additionally, JOPLEn (combined with a Dirty LASSO penalty) is an effective feature selection method for nonlinear prediction in multitask learning. Finally, we demonstrate the performance of JOPLEn on 153 regression and classification datasets and with a variety of penalties. JOPLEn leads to improved prediction performance relative to not only standard random forest and boosted tree ensembles, but also other methods for enhancing tree ensembles.

Despite implementing hundreds of strategies, cancer drug development suffers from a 95% failure rate over 30 years, with only 30% of approved cancer drugs extending patient survival beyond 2.5 months. Adding more criteria without eliminating nonessential ones is impractical and may fall into the "survivorship bias" trap. Machine learning (ML) models may enhance efficiency by saving time and cost. Yet, they may not improve success rate without identifying the root causes of failure. We propose a "STAR-guided ML system" (structure-tissue/cell selectivity-activity relationship) to enhance success rate and efficiency by addressing three overlooked interdependent factors: potency/specificity to the on/off-targets determining efficacy in tumors at clinical doses, on/off-target-driven tissue/cell selectivity influencing adverse effects in the normal organs at clinical doses, and optimal clinical doses balancing efficacy/safety as determined by potency/specificity and tissue/cell selectivity. STAR-guided ML models can directly predict clinical dose/efficacy/safety from five features to design/select the best drugs, enhancing success and efficiency of cancer drug development.

Abstract Motivation Understanding the mechanisms and structural mappings between molecules and pathway classes are critical for design of reaction predictors for synthesizing new molecules. This article studies the problem of prediction of classes of metabolic pathways (series of chemical reactions occurring within a cell) in which a given biochemical compound participates. We apply a hybrid machine learning approach consisting of graph convolutional networks used to extract molecular shape features as input to a random forest classifier. In contrast to previously applied machine learning methods for this problem, our framework automatically extracts relevant shape features directly from input SMILES representations, which are atom-bond specifications of chemical structures composing the molecules. Results Our method is capable of correctly predicting the respective metabolic pathway class of 95.16% of tested compounds, whereas competing methods only achieve an accuracy of 84.92% or less. Furthermore, our framework extends to the task of classification of compounds having mixed membership in multiple pathway classes. Our prediction accuracy for this multi-label task is 95.62%. We analyze the relative importance of various global physicochemical features to the pathway class prediction problem and show that simple linear/logistic regression models can predict the values of these global features from the shape features extracted using our framework. Availability and implementation https://github.com/baranwa2/MetabolicPathwayPrediction.

Tree ensembles achieve state-of-the-art performance on numerous prediction tasks. We propose $\textbf{J}$oint $\textbf{O}$ptimization of $\textbf{P}$iecewise $\textbf{L}$inear $\textbf{En}$sembles (JOPLEn), which jointly fits piecewise linear models at all leaf nodes of an existing tree ensemble. In addition to enhancing the ensemble expressiveness, JOPLEn allows several common penalties, including sparsity-promoting and subspace-norms, to be applied to nonlinear prediction. For example, JOPLEn with a nuclear norm penalty learns subspace-aligned functions. Additionally, JOPLEn (combined with a Dirty LASSO penalty) is an effective feature selection method for nonlinear prediction in multitask learning. Finally, we demonstrate the performance of JOPLEn on 153 regression and classification datasets and with a variety of penalties. JOPLEn leads to improved prediction performance relative to not only standard random forest and boosted tree ensembles, but also other methods for enhancing tree ensembles.

Extracting meaningful features from complex, high-dimensional datasets across scientific domains remains challenging. Current methods often struggle with scalability, limiting their applicability to large datasets, or make restrictive assumptions about feature-property relationships, hindering their ability to capture complex interactions. BoUTS's general and scalable feature selection algorithm surpasses these limitations to identify both universal features relevant to all datasets and task-specific features predictive for specific subsets. Evaluated on seven diverse chemical regression datasets, BoUTS achieves state-of-the-art feature sparsity while maintaining prediction accuracy comparable to specialized methods. Notably, BoUTS's universal features enable domain-specific knowledge transfer between datasets, and suggest deep connections in seemingly-disparate chemical datasets. We expect these results to have important repercussions in manually-guided inverse problems. Beyond its current application, BoUTS holds immense potential for elucidating data-poor systems by leveraging information from similar data-rich systems. BoUTS represents a significant leap in cross-domain feature selection, potentially leading to advancements in various scientific fields.

Polycyclic aromatic hydrocarbons are the main precursors to soot particles in combustion systems. A lack of direct experimental evidence has led to controversial theoretical explanations for the transition from gas-phase species to organic soot clusters. This work focuses on sampling infant soot particles from well-defined flames followed by analysis using state-of-the-art mass spectrometry. We found that PAH molecules present in soot particles are all stabilomers. Kinetic Monte Carlo simulations and thermodynamic stability calculations further identify the detected PAHs as peri-condensed and without aliphatic chains. Van der Waals forces can easily link PAHs of such size and shape to form PAH dimers and larger clusters under the specified flame conditions. Our results provide direct experimental evidence that soot inception is initiated by a physical process under typical flame conditions. This work improves our understanding of aerosol particulates, which has implications for their environmental and climate change impacts.

Finite-size effects in the static structure factor S(k) are analyzed for an amorphous substance. As the number of particles is reduced, S(0) increases greatly, up to an order of magnitude. Meanwhile, there is a decrease in the height of the first peak S_{peak}. These finite-size effects are modeled accurately by the Binder formula for S(0) and our empirical formula for S_{peak}. Procedures are suggested to correct for finite-size effects in S(k) data and in the hyperuniformity index H≡S(0)/S_{peak}. These principles generally apply to S(k) obtained from particle positions in noncrystalline substances. The amorphous substance we simulate is a two-dimensional liquid, with a soft Yukawa interaction modeling a dusty plasma experiment.

<p>A diverse range of polycyclic <a href="https://www.sciencedirect.com/topics/chemical-engineering/aromatic-compound" target="_blank">aromatic compounds</a> (PACs) is thought to exist in flame environments before and during soot inception. This work seeks to develop a machine learning (ML)-based soot inception model that considers detailed and diverse PAC properties such as <a href="https://www.sciencedirect.com/topics/chemical-engineering/oxygenation" target="_blank">oxygenation</a>, aliphatic content, radical character, size, and shape. To this end, temporal rates of change of PAC properties were computed by the stochastic modelling code SNapS2 and used as input to an ML model that predicts soot inception rate. The model is trained using experimentally-derived soot inception rates for three atmospheric pressure laminar premixed ethylene/air flames. An ML model (kernel ridge regression with a linear kernel) was developed to predict the soot inception rate in the three <a href="https://www.sciencedirect.com/topics/engineering/premixed-flame" target="_blank">premixed flames</a>. The soot inception rate predictions from this SNapS2-informed ML model outperformed the predictions from both the advanced soot <a href="https://www.sciencedirect.com/topics/engineering/computational-fluid-dynamic-modeling" target="_blank">modelling CFD</a> code CoFlame and an ML model which used CFD-determined inputs (temperature and species concentrations). The final model had an R^2 value of approximately 0.71 and a <a href="https://www.sciencedirect.com/topics/engineering/mean-absolute-error" target="_blank">mean absolute error</a> approximately 25% of the target values. The performance of the SNapS2-informed model suggests that detailed PAC properties are important to consider in inception modelling. While expanding this approach to other types of flames and fuels is crucial for future improvement to the model’s accuracy and generality, this methodology provides a successful framework for the current system. The success of this method demonstrates that ML can offer improvements in accuracy compared to current <a href="https://www.sciencedirect.com/topics/engineering/computational-fluid-dynamics" target="_blank">CFD</a> inception models and the highlights the potential for ML in soot predictions.</p>

<p>A diverse range of polycyclic <a href="https://www.sciencedirect.com/topics/chemical-engineering/aromatic-compound" target="_blank">aromatic compounds</a> (PACs) is thought to exist in flame environments before and during soot inception. This work seeks to develop a machine learning (ML)-based soot inception model that considers detailed and diverse PAC properties such as <a href="https://www.sciencedirect.com/topics/chemical-engineering/oxygenation" target="_blank">oxygenation</a>, aliphatic content, radical character, size, and shape. To this end, temporal rates of change of PAC properties were computed by the stochastic modelling code SNapS2 and used as input to an ML model that predicts soot inception rate. The model is trained using experimentally-derived soot inception rates for three atmospheric pressure laminar premixed ethylene/air flames. An ML model (kernel ridge regression with a linear kernel) was developed to predict the soot inception rate in the three <a href="https://www.sciencedirect.com/topics/engineering/premixed-flame" target="_blank">premixed flames</a>. The soot inception rate predictions from this SNapS2-informed ML model outperformed the predictions from both the advanced soot <a href="https://www.sciencedirect.com/topics/engineering/computational-fluid-dynamic-modeling" target="_blank">modelling CFD</a> code CoFlame and an ML model which used CFD-determined inputs (temperature and species concentrations). The final model had an R^2 value of approximately 0.71 and a <a href="https://www.sciencedirect.com/topics/engineering/mean-absolute-error" target="_blank">mean absolute error</a> approximately 25% of the target values. The performance of the SNapS2-informed model suggests that detailed PAC properties are important to consider in inception modelling. While expanding this approach to other types of flames and fuels is crucial for future improvement to the model’s accuracy and generality, this methodology provides a successful framework for the current system. The success of this method demonstrates that ML can offer improvements in accuracy compared to current <a href="https://www.sciencedirect.com/topics/engineering/computational-fluid-dynamics" target="_blank">CFD</a> inception models and the highlights the potential for ML in soot predictions.</p>

Amyloid nanofibers are abundant in microorganisms and are integral components of many biofilms, serving various purposes, from virulent to structural. Nonetheless, the precise characterization of bacterial amyloid nanofibers has been elusive, with incomplete and contradicting results. The present work focuses on the molecular details and characteristics of PSMα1-derived functional amyloids present in Staphylococcus aureus biofilms, using a combination of computational and experimental techniques, to develop a model that can aid the design of compounds to control amyloid formation. Results from molecular dynamics simulations, guided and supported by spectroscopy and microscopy, show that PSMα1 amyloid nanofibers present a helical structure formed by two protofilaments, have an average diameter of about 12 nm, and adopt a left-handed helicity with a periodicity of approximately 72 nm. The chirality of the self-assembled nanofibers, an intrinsic geometric property of its constituent peptides, is central to determining the fibers’ lateral growth.

A growing body of work has linked key biological activities to the mechanical properties of cellular membranes, and as a means of identification. Here, we present a computational approach to simulate and compare the vibrational spectra in the low-THz region for mammalian and bacterial membranes, investigating the effect of membrane asymmetry and composition, as well as the conserved frequencies of a specific cell. We find that asymmetry does not impact the vibrational spectra, and the impact of sterols depends on the mobility of the components of the membrane. We demonstrate that vibrational spectra can be used to distinguish between membranes and, therefore, could be used in identification of different organisms. The method presented, here, can be immediately extended to other biological structures (e.g., amyloid fibers, polysaccharides, and protein-ligand structures) in order to fingerprint and understand vibrations of numerous biologically-relevant nanoscale structures.

A growing body of work has linked key biological activities to the mechanical properties of cellular membranes, and as a means of identification. Here, we present a computational approach to simulate and compare the vibrational spectra in the low-THz region for mammalian and bacterial membranes, investigating the effect of membrane asymmetry and composition, as well as the conserved frequencies of a specific cell. We find that asymmetry does not impact the vibrational spectra, and the impact of sterols depends on the mobility of the components of the membrane. We demonstrate that vibrational spectra can be used to distinguish between membranes and, therefore, could be used in identification of different organisms. The method presented, here, can be immediately extended to other biological structures (e.g., amyloid fibers, polysaccharides, and protein-ligand structures) in order to fingerprint and understand vibrations of numerous biologically-relevant nanoscale structures.