Modeling transcriptional regulation of the cell cycle using a novel cybernetic-inspired approach.

Quantitative understanding of cellular processes, such as cell cycle and differentiation, is impeded by various forms of complexity ranging from myriad molecular players and their multilevel regulatory interactions, cellular evolution with multiple intermediate stages, lack of elucidation of cause-effect relationships among the many system players, and the computational complexity associated with the profusion of variables and parameters. In this paper, we present a modeling framework based on the cybernetic concept that biological regulation is inspired by objectives embedding rational strategies for dimension reduction, process stage specification through the system dynamics, and innovative causal association of regulatory events with the ability to predict the evolution of the dynamical system. The elementary step of the modeling strategy involves stage-specific objective functions that are computationally determined from experiments, augmented with dynamical network computations involving endpoint objective functions, mutual information, change-point detection, and maximal clique centrality. We demonstrate the power of the method through application to the mammalian cell cycle, which involves thousands of biomolecules engaged in signaling, transcription, and regulation. Starting with a fine-grained transcriptional description obtained from RNA sequencing measurements, we develop an initial model, which is then dynamically modeled using the cybernetic-inspired method, based on the strategies described above. The cybernetic-inspired method is able to distill the most significant interactions from a multitude of possibilities. In addition to capturing the complexity of regulatory processes in a mechanistically causal and stage-specific manner, we identify the functional network modules, including novel cell cycle stages. Our model is able to predict future cell cycles consistent with experimental measurements. We posit that this innovative framework has the promise to extend to the dynamics of other biological processes, with a potential to provide novel mechanistic insights.

Leveraging mathematical modeling to analyze nonadherence for hydroxyurea therapy in sickle cell disease.

Nonadherence is common in individuals with sickle cell disease (SCD) on hydroxyurea therapy and can be observed with waning improvements in hematologic parameters or biomarkers like mean cell volume and fetal hemoglobin level over time. We modeled the impact of hydroxyurea nonadherence on longitudinal biomarker profiles. We estimated the potential nonadherent days in individuals exhibiting a drop in biomarker levels by modifying the dosing profile using a probabilistic approach. Incorporating additional nonadherence using our approach besides existing ones in the dosing profile improves the model fits. We also studied how different patterns in adherence give rise to various physiological profiles of biomarkers. The key finding is consecutive days of nonadherence are less favorable than when nonadherence is interspersed. These findings improve our understanding of nonadherence and how appropriate intervention strategies can be applied for individuals with SCD susceptible to the severe impacts of nonadherence.

A comprehensive modeling approach for polymorph selection in Lennard-Jones crystallization.

Computational predictions of the polymorphic outcomes of a crystallization process, referred to as polymorph selection, can accelerate the process development for manufacturing solid products with targeted properties. Polymorph selection requires understanding the interplay between the thermodynamic and kinetic factors that drive nucleation. Moreover, post-nucleation events, such as crystal growth and polymorphic transformation, can affect the resulting crystal structures. Here, the nucleation kinetics of the Lennard-Jones (LJ) fluid from the melt is investigated with a focus on the competition between FCC and HCP crystal structures. Both molecular dynamics (MD) simulations and 2D free energy calculations reveal that polymorph selection occurs not during nucleation but when the cluster sizes exceed the critical cluster size. This result contrasts with the classical nucleation mechanism, where each polymorph is assumed to nucleate independently as an ideal bulk-like cluster, comprised only of its given structure. Using the 2D free energy surface and the MD simulation-derived diffusion coefficients, a structure-dependent nucleation rate is estimated, which agrees with the rate obtained from brute force MD simulations. Furthermore, a comprehensive population balance modeling (PBM) approach for polymorph selection is proposed. The PBM combines the calculated nucleation rate with post-nucleation kinetics while accounting for the structural changes of the clusters after nucleation. When applied to the LJ system, the PBM predicts with high accuracy the polymorphic distribution found in a population of crystals generated from MD simulations. Due to the non-classical nucleation mechanism of the LJ system, post-nucleation kinetic events are crucial in determining the structures of the grown crystals.

Modeling transcriptional regulation of the cell cycle using a novel cybernetic-inspired approach.

Quantitative understanding of cellular processes, such as cell cycle and differentiation, is impeded by various forms of complexity ranging from myriad molecular players and their multilevel regulatory interactions, cellular evolution with multiple intermediate stages, lack of elucidation of cause-effect relationships among the many system players, and the computational complexity associated with the profusion of variables and parameters. In this paper, we present an elegant modeling framework based on the cybernetic concept that biological regulation is inspired by objectives embedding entirely novel strategies for dimension reduction, process stage specification through the system dynamics, and innovative causal association of regulatory events with the ability to predict the evolution of the dynamical system. The elementary step of the modeling strategy involves stage-specific objective functions that are computationally-determined from experiments, augmented with dynamical network computations involving end point objective functions, mutual information, change point detection, and maximal clique centrality. We demonstrate the power of the method through application to the mammalian cell cycle, which involves thousands of biomolecules engaged in signaling, transcription, and regulation. Starting with a fine-grained transcriptional description obtained from RNA sequencing measurements, we develop an initial model, which is then dynamically modeled using the cybernetic-inspired method (CIM), utilizing the strategies described above. The CIM is able to distill the most significant interactions from a multitude of possibilities. In addition to capturing the complexity of regulatory processes in a mechanistically causal and stage-specific manner, we identify the functional network modules, including novel cell cycle stages. Our model is able to predict future cell cycles consistent with experimental measurements. We posit that this state-of-the-art framework has the promise to extend to the dynamics of other biological processes, with a potential to provide novel mechanistic insights. STATEMENT OF SIGNIFICANCE: Cellular processes like cell cycle are overly complex, involving multiple players interacting at multiple levels, and explicit modeling of such systems is challenging. The availability of longitudinal RNA measurements provides an opportunity to "reverse-engineer" for novel regulatory models. We develop a novel framework, inspired using goal-oriented cybernetic model, to implicitly model transcriptional regulation by constraining the system using inferred temporal goals. A preliminary causal network based on information-theory is used as a starting point, and our framework is used to distill the network to temporally-based networks containing essential molecular players. The strength of this approach is its ability to dynamically model the RNA temporal measurements. The approach developed paves the way for inferring regulatory processes in many complex cellular processes.

Modeling COVID-19 transmission between age groups in the United States considering virus mutations, vaccinations, and reinfection.

The in-depth understanding of the dynamics of COVID-19 transmission among different age groups is of great interest for governments and health authorities so that strategies can be devised to reduce the pandemic's detrimental effects. We developed the SIRDV-Virulence (Susceptible-Infected-Recovered-Dead-Vaccinated-Virulence) epidemiological model based on a population balance equation to study the effects virus mutants, vaccination strategies, 'Anti/Non Vaxxer' proportions, and reinfection rates to provide methods to mitigate COVID-19 transmission among the United States population. Based on publicly available data, we obtain the key parameters governing the spread of the pandemic. The results show that a large fraction of infected cases comes from the adult and children populations in the presence of a highly infectious COVID-19 mutant. Given the situation at the end of July 2021, the results show that prioritizing children and adult vaccinations over that of seniors can contain the spread of the active cases, thereby preventing the healthcare system from being overwhelmed and minimizing subsequent deaths. The model suggests that the only option to curb the effects of this pandemic is to reduce the population of unvaccinated individuals. A higher fraction of 'Anti/Non-vaxxers' and a higher reinfection rate can both independently lead to the resurgence of the pandemic.

Mathematical Modeling of Hydroxyurea Therapy in Individuals with Sickle Cell Disease.

Sickle cell disease (SCD) is a chronic hemolytic anemia affecting millions worldwide with acute and chronic clinical manifestations and early mortality. While hydroxyurea (HU) and other treatment strategies managed to ameliorate disease severity, high inter-individual variability in clinical response and a lack of an ability to predict those variations need to be addressed to maximize the clinical efficacy of HU. We developed pharmacokinetics (PK) and pharmacodynamics (PD) models to study the dosing, efficacy, toxicity, and clinical response of HU treatment in more than eighty children with SCD. The clinical PK parameters were used to model the HU plasma concentration for a 24 h period, and the estimated daily average HU plasma concentration was used as an input to our PD models with approximately 1 to 9 years of data connecting drug exposure with drug response. We modeled the biomarkers mean cell volume and fetal hemoglobin to study treatment efficacy. For myelosuppression, we modeled red blood cells and absolute neutrophil count. Our models provided excellent fits for individuals with known or correctly inferred adherence. Our models can be used to determine the optimal dosing regimens and study the effect of non-adherence on HU-treated individuals.

Brownian bridges for stochastic chemical processes-An approximation method based on the asymptotic behavior of the backward Fokker-Planck equation.

A Brownian bridge is a continuous random walk conditioned to end in a given region by adding an effective drift to guide paths toward the desired region of phase space. This idea has many applications in chemical science where one wants to control the endpoint of a stochastic process-e.g., polymer physics, chemical reaction pathways, heat/mass transfer, and Brownian dynamics simulations. Despite its broad applicability, the biggest limitation of the Brownian bridge technique is that it is often difficult to determine the effective drift as it comes from a solution of a Backward Fokker-Planck (BFP) equation that is infeasible to compute for complex or high-dimensional systems. This paper introduces a fast approximation method to generate a Brownian bridge process without solving the BFP equation explicitly. Specifically, this paper uses the asymptotic properties of the BFP equation to generate an approximate drift and determine ways to correct (i.e., re-weight) any errors incurred from this approximation. Because such a procedure avoids the solution of the BFP equation, we show that it drastically accelerates the generation of conditioned random walks. We also show that this approach offers reasonable improvement compared to other sampling approaches using simple bias potentials.

An investigation of the kinetics and thermodynamics of NaCl nucleation through composite clusters.

Having a good understanding of nucleation is critical for the control of many important processes, such as polymorph selection during crystallization. However, a complete picture of the molecular-level mechanisms of nucleation remains elusive. In this work, we take an in-depth look at the NaCl homogeneous nucleation mechanism through thermodynamics. Distinguished from the classical nucleation theory, we calculate the free energy of nucleation as a function of two nucleus size coordinates: crystalline and amorphous cluster sizes. The free energy surface reveals a thermodynamic preference for a nonclassical mechanism of nucleation through a composite cluster, where the crystalline nucleus is surrounded by an amorphous layer. The thickness of the amorphous layer increases with an increase in supersaturation. The computed free energy landscape agrees well with the composite cluster-free energy model, through which phase specific thermodynamic properties are evaluated. As the supersaturation increases, there is a change in stability of the amorphous phase relative to the solution phase, resulting in a change from one-step to two-step mechanism, seen clearly from the free energy profile along the minimum free energy path crossing the transition curve. By obtaining phase-specific diffusion coefficients, we construct the full mesoscopic model and present a clear roadmap for NaCl nucleation.

Examining Sodium and Potassium Channel Conductances Involved in Hyperexcitability of Chemotherapy-Induced Peripheral Neuropathy: A Mathematical and Cell Culture-Based Study.

Chemotherapy-induced peripheral neuropathy (CIPN) is a prevalent, painful side effect which arises due to a number of chemotherapy agents. CIPN can have a prolonged effect on quality of life. Chemotherapy treatment is often reduced or stopped altogether because of the severe pain. Currently, there are no FDA-approved treatments for CIPN partially due to its complex pathogenesis in multiple pathways involving a variety of channels, specifically, voltage-gated ion channels. One aspect of neuropathic pain in vitro is hyperexcitability in dorsal root ganglia (DRG) peripheral sensory neurons. Our study employs bifurcation theory to investigate the role of voltage-gated ion channels in inducing hyperexcitability as a consequence of spontaneous firing due to the common chemotherapy agent paclitaxel. Our mathematical investigation of a reductionist DRG neuron model comprised of sodium channel Nav1.7, sodium channel Nav1.8, delayed rectifier potassium channel, A-type transient potassium channel, and a leak channel suggests that Nav1.8 and delayed rectifier potassium channel conductances are critical for hyperexcitability of small DRG neurons. Introducing paclitaxel into the model, our bifurcation analysis predicts that hyperexcitability is highest for a medium dose of paclitaxel, which is supported by multi-electrode array (MEA) recordings. Furthermore, our findings using MEA reveal that Nav1.8 blocker A-803467 and delayed rectifier potassium enhancer L-alpha-phosphatidyl-D-myo-inositol 4,5-diphosphate, dioctanoyl (PIP2) can reduce paclitaxel-induced hyperexcitability of DRG neurons. Our approach can be readily extended and used to investigate various other contributors of hyperexcitability in CIPN.

Computational analysis of a 9D model for a small DRG neuron.

Small dorsal root ganglion (DRG) neurons are primary nociceptors which are responsible for sensing pain. Elucidation of their dynamics is essential for understanding and controlling pain. To this end, we present a numerical bifurcation analysis of a small DRG neuron model in this paper. The model is of Hodgkin-Huxley type and has 9 state variables. It consists of a Nav1.7 and a Nav1.8 sodium channel, a leak channel, a delayed rectifier potassium, and an A-type transient potassium channel. The dynamics of this model strongly depend on the maximal conductances of the voltage-gated ion channels and the external current, which can be adjusted experimentally. We show that the neuron dynamics are most sensitive to the Nav1.8 channel maximal conductance ([Formula: see text]). Numerical bifurcation analysis shows that depending on [Formula: see text] and the external current, different parameter regions can be identified with stable steady states, periodic firing of action potentials, mixed-mode oscillations (MMOs), and bistability between stable steady states and stable periodic firing of action potentials. We illustrate and discuss the transitions between these different regimes. We further analyze the behavior of MMOs. As the external current is decreased, we find that MMOs appear after a cyclic limit point. Within this region, bifurcation analysis shows a sequence of isolated periodic solution branches with one large action potential and a number of small amplitude peaks per period. For decreasing external current, the number of small amplitude peaks is increasing and the distance between the large amplitude action potentials is growing, finally tending to infinity and thereby leading to a stable steady state. A closer inspection reveals more complex concatenated MMOs in between these periodic MMO branches, forming Farey sequences. Lastly, we also find small solution windows with aperiodic oscillations which seem to be chaotic. The dynamical patterns found here-as consequences of bifurcation points regulated by different parameters-have potential translational significance as repetitive firing of action potentials imply pain of some form and intensity; manipulating these patterns by regulating the different parameters could aid in investigating pain dynamics.

Exact sampling of polymer conformations using Brownian bridges.

The equilibrium conformation of a polymer molecule in an external field is often used in field theories to calculate macroscopic polymer properties of melts and solutions. We use a mathematical method called a Brownian bridge to exactly sample continuous polymer chains to end in a given state. We show that one can systematically develop such processes to sample specific polymer topologies, to confine polymers in a given geometry for its entire path, to efficiently generate high-probability conformations by excluding small Boltzmann weights, or to simulate rare events in a rugged energy landscape. This formalism can improve the polymer sampling efficiency significantly compared to traditional methods (e.g., Monte Carlo or Rosenbluth).

A Metabolomics Approach for Early Prediction of Vincristine-Induced Peripheral Neuropathy.

Vincristine is a core chemotherapeutic drug administered to pediatric acute lymphoblastic leukemia patients. Despite its efficacy in treating leukemia, it can lead to severe peripheral neuropathy in a subgroup of the patients. Peripheral neuropathy is a debilitating and painful side-effect that can severely impact an individual's quality of life. Currently, there are no established predictors of peripheral neuropathy incidence during the early stage of chemotherapeutic treatment. As a result, patients who are not susceptible to peripheral neuropathy may receive sub-therapeutic treatment due to an empirical upper cap on the dose, while others may experience severe neuropathy at the same dose. Contrary to previous genomics based approaches, we employed a metabolomics approach to identify small sets of metabolites that can be used to predict a patient's susceptibility to peripheral neuropathy at different time points during the treatment. Using those identified metabolites, we developed a novel strategy to predict peripheral neuropathy and subsequently adjust the vincristine dose accordingly. In accordance with this novel strategy, we created a free user-friendly tool, VIPNp, for physicians to easily implement our prediction strategy. Our results showed that focusing on metabolites, which encompasses both genotypic and phenotypic variations, can enable early prediction of peripheral neuropathy in pediatric leukemia patients.

A model to rate strategies for managing disease due to COVID-19 infection.

Considering looming fatality and economic recession, effective policy making based on ongoing COVID-19 pandemic is an urgent and standing issue. Numerous issues for controlling infection have arisen from public discussion led by medical professionals. Yet understanding of these factors has been necessarily qualitative and control measures to correct unfavorable trends specific to an infection area have been lacking. The logical implement for control is a large scale stochastic model with countless parameters lacking robustness and requiring enormous data. This paper presents a remedy for this vexing problem by proposing an alternative approach. Machine learning has come to play a widely circulated role in the study of complex data in recent times. We demonstrate that when machine learning is employed together with the mechanistic framework of a mathematical model, there can be a considerably enhanced understanding of complex systems. A mathematical model describing the viral infection dynamics reveals two transmissibility parameters influenced by the management strategies in the area for the control of the current pandemic. Both parameters readily yield the peak infection rate and means for flattening the curve, which is correlated to different management strategies by employing machine learning, enabling comparison of different strategies and suggesting timely alterations. Treatment of population data with the model shows that restricted non-essential business closure, school closing and strictures on mass gathering influence the spread of infection. While a rational strategy for initiation of an economic reboot would call for a wider perspective of the local economics, the model can speculate on its timing based on the status of the infection as reflected by its potential for an unacceptably renewed viral onslaught.

Erratum: Addition to "Revised Formulation of Fick's, Fourier's, and Newton's Laws for Spatially Varying Linear Transport Coefficients".

[This corrects the article DOI: 10.1021/acsomega.9b00736.].

Revised Formulation of Fick's, Fourier's, and Newton's Laws for Spatially Varying Linear Transport Coefficients.

We argue that for situations involving spatially varying linear transport coefficients (diffusivities, thermal conductivities, and viscosities), the original Fick's, Fourier's, and Newton's law equations should be modified to place the diffusivity, thermal conductivity, and viscosity inside the derivative operator, that is, in one-dimensional rectilinear situations, , , and . We present simple derivations of these formulas in which diffusive mass transfer, conductive heat transfer, and molecular momentum transfer processes are described using lattice random walk models. We also present simple examples demonstrating how these modifications affect calculations.

Consistency of cybernetic variables with gene expression profiles: A more rigorous test.

Diauxic growth of Escherichia coli is driven by a host of internal, complex regulatory actions. In this classic scenario of cellular control, the cell employs a rational algorithm to modulate its metabolism in a competitive fashion. Cybernetic models of metabolism, whose development now spans three decades, were first formulated to describe regulation of cells in complex, multi-substrate environments. They modeled this scenario using the hypothesis that the formation of the enzymatic machinery is regulated to maximize a return on investment. While this assumption is made on the basis of logical arguments rooted in evolutionary principles, little effort has been taken to validate if enzymes are truly synthesized in the same fashion that is predicted by cybernetic variables. This work revisits the original cybernetic models describing diauxic growth and compares their predictions of enzyme synthesis control with time series gene expression data in microarray and qRT-PCR formats. Three separate studies are made for two different strains of E. coli. The first is for the growth of E. coli BW25113 on a mixture of glucose and acetate, whose gene expression changes are metered by microarray. Another is also for the sequential consumption of glucose and acetate but involves strain MG1655 and employs qRT-PCR. The final is for E. coli MG1655 on glucose and lactose. By demonstrating how cybernetic variables for induced enzyme synthesis mimic the behavior of transcriptional data, a strong argument for using cybernetic models is made. © 2018 American Institute of Chemical Engineers Biotechnol. Prog., 34:858-867, 2018.

Insulin exits skeletal muscle capillaries by fluid-phase transport.

Before insulin can stimulate myocytes to take up glucose, it must first move from the circulation to the interstitial space. The continuous endothelium of skeletal muscle (SkM) capillaries restricts insulin's access to myocytes. The mechanism by which insulin crosses this continuous endothelium is critical to understand insulin action and insulin resistance; however, methodological obstacles have limited understanding of endothelial insulin transport in vivo. Here, we present an intravital microscopy technique to measure the rate of insulin efflux across the endothelium of SkM capillaries. This method involves development of a fully bioactive, fluorescent insulin probe, a gastrocnemius preparation for intravital microscopy, an automated vascular segmentation algorithm, and the use of mathematical models to estimate endothelial transport parameters. We combined direct visualization of insulin efflux from SkM capillaries with modeling of insulin efflux kinetics to identify fluid-phase transport as the major mode of transendothelial insulin efflux in mice. Model-independent experiments demonstrating that insulin movement is neither saturable nor affected by insulin receptor antagonism supported this result. Our finding that insulin enters the SkM interstitium by fluid-phase transport may have implications in the pathophysiology of SkM insulin resistance as well as in the treatment of diabetes with various insulin analogs.

Sequential computation of elementary modes and minimal cut sets in genome-scale metabolic networks using alternate integer linear programming.

MOTIVATION: Elementary (flux) modes (EMs) have served as a valuable tool for investigating structural and functional properties of metabolic networks. Identification of the full set of EMs in genome-scale networks remains challenging due to combinatorial explosion of EMs in complex networks. It is often, however, that only a small subset of relevant EMs needs to be known, for which optimization-based sequential computation is a useful alternative. Most of the currently available methods along this line are based on the iterative use of mixed integer linear programming (MILP), the effectiveness of which significantly deteriorates as the number of iterations builds up. To alleviate the computational burden associated with the MILP implementation, we here present a novel optimization algorithm termed alternate integer linear programming (AILP). RESULTS: Our algorithm was designed to iteratively solve a pair of integer programming (IP) and linear programming (LP) to compute EMs in a sequential manner. In each step, the IP identifies a minimal subset of reactions, the deletion of which disables all previously identified EMs. Thus, a subsequent LP solution subject to this reaction deletion constraint becomes a distinct EM. In cases where no feasible LP solution is available, IP-derived reaction deletion sets represent minimal cut sets (MCSs). Despite the additional computation of MCSs, AILP achieved significant time reduction in computing EMs by orders of magnitude. The proposed AILP algorithm not only offers a computational advantage in the EM analysis of genome-scale networks, but also improves the understanding of the linkage between EMs and MCSs. AVAILABILITY AND IMPLEMENTATION: The software is implemented in Matlab, and is provided as supplementary information . CONTACT: hyunseob.song@pnnl.gov. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Nanocrystal Dissolution Kinetics and Solubility Increase Prediction from Molecular Dynamics: The Case of α-, ß-, and γ-Glycine.

Nanocrystals are receiving increased attention for pharmaceutical applications due to their enhanced solubility relative to their micron-sized counterpart and, in turn, potentially increased bioavailability. In this work, a computational method is proposed to predict the following: (1) polymorph specific dissolution kinetics and (2) the multiplicative increase in the polymorph specific nanocrystal solubility relative to the bulk solubility. The method uses a combination of molecular dynamics and a parametric particle size dependent mass transfer model. The method is demonstrated using a case study of α-, ß-, and γ-glycine. It is shown that only the α-glycine form is predicted to have an increasing dissolution rate with decreasing particle size over the range of particle sizes simulated. On the contrary, γ-glycine shows a monotonically increasing dissolution rate with increasing particle size and dissolves at a rate 1.5 to 2 times larger than α- or ß-glycine. The accelerated dissolution rate of γ-glycine relative to the other two polymorphs correlates directly with the interfacial energy ranking of γ > ß > α obtained from the dissolution simulations, where γ- is predicted to have an interfacial energy roughly four times larger than either α- or ß-glycine. From the interfacial energies, α- and ß-glycine nanoparticles were predicted to experience modest solubility increases of up to 1.4 and 1.8 times the bulk solubility, where as γ-glycine showed upward of an 8 times amplification in the solubility. These MD simulations represent a first attempt at a computational (pre)screening method for the rational design of experiments for future engineering of nanocrystal API formulations.

Solubility curves and nucleation rates from molecular dynamics for polymorph prediction - moving beyond lattice energy minimization.

Current polymorph prediction methods, known as lattice energy minimization, seek to determine the crystal lattice with the lowest potential energy, rendering it unable to predict solvent dependent metastable form crystallization. Facilitated by embarrassingly parallel, multiple replica, large-scale molecular dynamics simulations, we report on a new method concerned with predicting crystal structures using the kinetics and solubility of the low energy polymorphs predicted by lattice energy minimization. The proposed molecular dynamics simulation methodology provides several new predictions to the field of crystallization. (1) The methodology is shown to correctly predict the kinetic preference for ß-glycine nucleation in water relative to α- and γ-glycine. (2) Analysis of nanocrystal melting temperatures show γ- nanocrystals have melting temperatures up to 20 K lower than either α- or ß-glycine. This provides a striking explanation of how an energetically unstable classical nucleation theory (CNT) transition state complex leads to kinetic inaccessibility of γ-glycine in water, despite being the thermodynamically preferred polymorph predicted by lattice energy minimization. (3) The methodology also predicts polymorph-specific solubility curves, where the α-glycine solubility curve is reproduced to within 19% error, over a 45 K temperature range, using nothing but atomistic-level information provided from nucleation simulations. (4) Finally, the methodology produces the correct solubility ranking of ß- > α-glycine. In this work, we demonstrate how the methodology supplements lattice energy minimization with molecular dynamics nucleation simulations to give the correct polymorph prediction, at different length scales, when lattice energy minimization alone would incorrectly predict the formation of γ-glycine in water from the ranking of lattice energies. Thus, lattice energy minimization optimization algorithms are supplemented with the necessary solvent/solute dependent solubility and nucleation kinetics of polymorphs to predict which structure will come out of solution, and not merely which structure has the most stable lattice energy.