Search

Proteins naturally occur in crowded cellular environments and interact with other proteins, nucleic acids, and organelles. Since most previous experimental protein structure determination techniques require that proteins occur in idealized, non-physiological environments, the effects of realistic cellular environments on protein structure are largely unexplored. Recently, Förster resonance energy transfer (FRET) has been shown to be an effective experimental method for investigating protein structure in vivo . Inter-residue distances measured in vivo can be incorporated as restraints in molecular dynamics (MD) simulations to model protein structural dynamics in vivo . Since most FRET studies only obtain inter-residue separations for a small number of amino acid pairs, it is important to determine the minimum number of restraints in the MD simulations that are required to achieve a given root-mean-square deviation (RMSD) from the experimental structural ensemble. Further, what is the optimal method for selecting these inter-residue restraints? Here, we implement several methods for selecting the most important FRET pairs and determine the number of pairs N r that are needed to induce conformational changes in proteins between two experimentally determined structures. We find that enforcing only a small fraction of restraints, N r / N ≲ 0.08 , where N is the number of amino acids, can induce the conformational changes. These results establish the efficacy of FRET-assisted MD simulations for atomic scale structural modeling of proteins in vivo ., Competing Interests: Conflict of interest statement: The authors declare no conflicts of interests.

Chromatin is a polymer complex of DNA and proteins that regulates gene expression. The three-dimensional (3D) structure and organization of chromatin controls DNA transcription and replication. High-throughput chromatin conformation capture techniques generate Hi-C maps that can provide insight into the 3D structure of chromatin. Hi-C maps can be represented as a symmetric matrix [Formula: see text], where each element represents the average contact probability or number of contacts between chromatin loci i and j. Previous studies have detected topologically associating domains (TADs), or self-interacting regions in [Formula: see text] within which the contact probability is greater than that outside the region. Many algorithms have been developed to identify TADs within Hi-C maps. However, most TAD identification algorithms are unable to identify nested or overlapping TADs and for a given Hi-C map there is significant variation in the location and number of TADs identified by different methods. We develop a novel method to identify TADs, KerTAD, using a kernel-based technique from computer vision and image processing that is able to accurately identify nested and overlapping TADs. We benchmark this method against state-of-the-art TAD identification methods on both synthetic and experimental data sets. We find that the new method consistently has higher true positive rates (TPR) and lower false discovery rates (FDR) than all tested methods for both synthetic and manually annotated experimental Hi-C maps. The TPR for KerTAD is also largely insensitive to increasing noise and sparsity, in contrast to the other methods. We also find that KerTAD is consistent in the number and size of TADs identified across replicate experimental Hi-C maps for several organisms. Thus, KerTAD will improve automated TAD identification and enable researchers to better correlate changes in TADs to biological phenomena, such as enhancer-promoter interactions and disease states., Competing Interests: The authors have declared that no competing interests exist., (Copyright: © 2024 Maisuradze et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.)

Breast cancer invasion into adipose tissue strongly influences disease progression and metastasis. The degree of cancer cell invasion into adipose tissue depends on both biochemical signaling and the mechanical properties of cancer cells, adipocytes, and other key components of adipose tissue. We model breast cancer invasion into adipose tissue using discrete element method simulations of active, cohesive spherical particles (cancer cells) invading into confluent packings of deformable polyhedra (adipocytes). We quantify the degree of invasion by calculating the interfacial area A t between cancer cells and adipocytes. We determine the long-time value of A t vs the activity and strength of the cohesion between cancer cells, as well as the mechanical properties of the adipocytes and extracellular matrix in which adipocytes are embedded. We show that the degree of invasion collapses onto a master curve as a function of the dimensionless energy scale E c , which grows linearly with the cancer cell velocity persistence time and fluctuations, is inversely proportional to the system pressure, and is offset by the cancer cell cohesive energy. When E c > 1 , cancer cells will invade the adipose tissue, whereas for E c < 1 , cancer cells and adipocytes remain de-mixed. We also show that A t decreases when the adipocytes are constrained by the ECM by an amount that depends on the spatial heterogeneity of the adipose tissue., Competing Interests: The authors have no conflicts to disclose., (© 2024 Author(s).)

Swarming is a macroscopic phenomenon in which surface bacteria organize into a motile population. The flagellar motor that drives swarming in Pseudomonas aeruginosa is powered by stators MotAB and MotCD. Deletion of the MotCD stator eliminates swarming, whereas deletion of the MotAB stator enhances swarming. Interestingly, we measured a strongly asymmetric stator availability in the wild-type (WT) strain, with MotAB stators produced at an approximately 40-fold higher level than MotCD stators. However, utilization of MotCD stators in free swimming cells requires higher liquid viscosities, while MotAB stators are readily utilized at low viscosities. Importantly, we find that cells with MotCD stators are ~10× more likely to have an active motor compared to cells uses the MotAB stators. The spectrum of motility intermittency can either cooperatively shut down or promote flagellum motility in WT populations. In P. aeruginosa , transition from a static solid-like biofilm to a dynamic liquid-like swarm is not achieved at a single critical value of flagellum torque or stator fraction but is collectively controlled by diverse combinations of flagellum activities and motor intermittencies via dynamic stator utilization. Experimental and computational results indicate that the initiation or arrest of flagellum-driven swarming motility does not occur from individual fitness or motility performance but rather related to concepts from the "jamming transition" in active granular matter.IMPORTANCEIt is now known that there exist multifactorial influences on swarming motility for P. aeruginosa , but it is not clear precisely why stator selection in the flagellum motor is so important. We show differential production and utilization of the stators. Moreover, we find the unanticipated result that the two motor configurations have significantly different motor intermittencies: the fraction of flagellum-active cells in a population on average with MotCD is active ~10× more often than with MotAB. What emerges from this complex landscape of stator utilization and resultant motor output is an intrinsically heterogeneous population of motile cells. We show how consequences of stator recruitment led to swarming motility and how the stators potentially relate to surface sensing circuitry., Competing Interests: The authors declare no conflict of interest.

The mechanical properties of CuTi alloys have been characterized extensively through experimental studies. However, a detailed understanding of why the strength of Cu increases after a small fraction of Ti atoms are added to the alloy is still missing. In this work, we address this question using density functional theory (DFT) and molecular dynamics (MD) simulations with the modified embedded atom method (MEAM) interatomic potentials. First, we performed calculations of the uniaxial tension deformations of small bicrystalline Cu cells using DFT static simulations. We then carried out uniaxial tension deformations on much larger bicrystalline and polycrystalline Cu cells by using MEAM MD simulations. In bicrystalline Cu, the inclusion of Ti increases the grain boundary separation energy and the maximum tensile stress. The DFT calculations demonstrate that the increase in the tensile stress can be attributed to an increase in the local charge density arising from Ti. MEAM simulations in larger bicrystalline systems have shown that increasing the Ti concentration decreases the density of the stacking faults. This observation is enhanced in polycrystalline Cu, where the addition of Ti atoms, even at concentrations as low as 1.5 atomic (at.) %, increases the yield strength and elastic modulus of the material compared to pure Cu. Under uniaxial tensile loading, the addition of small amounts of Ti hinders the formation of partial Shockley dislocations in the grain boundaries of Cu, leading to a reduced level of local deformation. These results shed light on the role of Ti in determining the mechanical properties of polycrystalline Cu and enable the engineering of grain boundaries and the inclusion of Ti to improve degradation resistance., Competing Interests: The authors declare no competing financial interest., (© 2024 The Authors. Published by American Chemical Society.)

The spongy mesophyll is a complex, porous tissue found in plant leaves that enables carbon capture and provides mechanical stability. Unlike many other biological tissues, which remain confluent throughout development, the spongy mesophyll must develop from an initially confluent tissue into a tortuous network of cells with a large proportion of intercellular airspace. How the airspace in the spongy mesophyll develops while the tissue remains mechanically stable is unknown. Here, we use computer simulations of deformable polygons to develop a purely mechanical model for the development of the spongy mesophyll tissue. By stipulating that cell wall growth and remodelling occurs only near void space, our computational model is able to recapitulate spongy mesophyll development observed in Arabidopsis thaliana leaves. We find that robust generation of pore space in the spongy mesophyll requires a balance of cell growth, adhesion, stiffness and tissue pressure to ensure cell networks become porous yet maintain mechanical stability. The success of this mechanical model of morphogenesis suggests that simple physical principles can coordinate and drive the development of complex plant tissues like the spongy mesophyll.

The Integrated Graduate Program in Physical and Engineering Biology (IGPPEB) at Yale University brings together Ph.D. students from the physical, engineering, and biological sciences. The main goals of this program are for students to become comfortable working in an interdisciplinary and collaborative research environment and adept at communicating with scientists and nonscientists. To fill a student-identified learning gap in engaging in inclusive discussions, IGPPEB students developed a communication workshop to improve skills in visual engagement, citing specific content, constructive conversation entrances, and encouragement of peers. Based on short- and long-term assessment of the workshop, 100% of students reported that it should be offered to future cohorts and 63% of students perceived it to be personally helpful. Additionally, 92% of participants reported using one or more of the core skills beyond the course, with skills in "Encouraging peers" and "Constructive conversation entrances" rated the highest in perceived improvement. Based on the highest average rating of 76 ± 24 (on a scale of 0-100), students agreed that the workshop made them feel more welcome in the IGPPEB community. With a rating of 68 ± 13, they also agreed that the workshop had a positive impact on their graduate school experience. Participants provided suggestions for future improvements, such as increasing student involvement in leading discussions of course material. This study demonstrates that a student-led workshop can improve perceived discussion skills and build community across an interdisciplinary program in the sciences., Competing Interests: The authors have stated explicitly that there are no conflicts of interest in connection with this article., (©2022 The Authors FASEB BioAdvances published by The Federation of American Societies for Experimental Biology.)

Numerous studies have investigated the differences and similarities between protein structures determined by solution NMR spectroscopy and those determined by X-ray crystallography. A fundamental question is whether any observed differences are due to differing methodologies or to differences in the behavior of proteins in solution versus in the crystalline state. Here, we compare the properties of the hydrophobic cores of high-resolution protein crystal structures and those in NMR structures, determined using increasing numbers and types of restraints. Prior studies have reported that many NMR structures have denser cores compared with those of high-resolution X-ray crystal structures. Our current work investigates this result in more detail and finds that these NMR structures tend to violate basic features of protein stereochemistry, such as small non-bonded atomic overlaps and few Ramachandran and sidechain dihedral angle outliers. We find that NMR structures solved with more restraints, and which do not significantly violate stereochemistry, have hydrophobic cores that have a similar size and packing fraction as their counterparts determined by X-ray crystallography at high resolution. These results lead us to conclude that, at least regarding the core packing properties, high-quality structures determined by NMR and X-ray crystallography are the same, and the differences reported earlier are most likely a consequence of methodology, rather than fundamental differences between the protein in the two different environments., (© 2022 The Protein Society.)

Functional particles that respond to external stimuli are spurring technological evolution across various disciplines. While large-scale production of functional particles is needed for their use in real-life applications, precise control over particle shapes and directional properties has remained elusive for high-throughput processes. We developed a high-throughput emulsion-based process that exploits rapid vitrification of a thixotropic medium to manufacture diverse functional particles in large quantities. The vitrified medium renders stationary emulsion droplets that preserve their shape and size during solidification, and energetic fields can be applied to build programmed anisotropy into the particles. We showcase mass-production of several functional particles, including low-melting point metallic particles, self-propelling Janus particles, and unidirectionally-magnetized robotic particles, via this static-state particle fabrication process.

The ability to consistently distinguish real protein structures from computationally generated model decoys is not yet a solved problem. One route to distinguish real protein structures from decoys is to delineate the important physical features that specify a real protein. For example, it has long been appreciated that the hydrophobic cores of proteins contribute significantly to their stability. We used two sources to obtain datasets of decoys to compare with real protein structures: submissions to the biennial Critical Assessment of protein Structure Prediction competition, in which researchers attempt to predict the structure of a protein only knowing its amino acid sequence, and also decoys generated by 3DRobot, which have user-specified global root-mean-squared deviations from experimentally determined structures. Our analysis revealed that both sets of decoys possess cores that do not recapitulate the key features that define real protein cores. In particular, the model structures appear more densely packed (because of energetically unfavorable atomic overlaps), contain too few residues in the core, and have improper distributions of hydrophobic residues throughout the structure. Based on these observations, we developed a feed-forward neural network, which incorporates key physical features of protein cores, to predict how well a computational model recapitulates the real protein structure without knowledge of the structure of the target sequence. By identifying the important features of protein structure, our method is able to rank decoy structures with similar accuracy to that obtained by state-of-the-art methods that incorporate many additional features. The small number of physical features makes our model interpretable, emphasizing the importance of protein packing and hydrophobicity in protein structure prediction., (© 2020 The Protein Society.)

Embryonic organizers establish gradients of diffusible signaling molecules to pattern the surrounding cells. Here, we elucidate an additional mechanism of embryonic organizers that is a secondary consequence of morphogen signaling. Using pharmacological and localized transgenic perturbations, 4D imaging of the zebrafish embryo, systematic analysis of cell motion, and computational modeling, we find that the vertebrate tail organizer orchestrates morphogenesis over distances beyond the range of morphogen signaling. The organizer regulates the rate and coherence of cell motion in the elongating embryo using mechanical information that is transmitted via relay between neighboring cells. This mechanism is similar to a pressure front in granular media and other jammed systems, but in the embryo the mechanical information emerges from self-propelled cell movement and not force transfer between cells. The propagation likely relies upon local biochemical signaling that affects cell contractility, cell adhesion, and/or cell polarity but is independent of transcription and translation., (Copyright © 2019 Elsevier Inc. All rights reserved.)

While common growth models assume a structure-less liquid composed of atomic flow units, structural ordering has been shown in liquid metals. Here, we conduct in situ transmission electron microscopy crystallization experiments on metallic glass nanorods, and show that structural ordering strongly affects crystal growth and is controlled by nanorod thermal history. Direct visualization reveals structural ordering as densely populated small clusters in a nanorod heated from the glass state, and similar behavior is found in molecular dynamics simulations of model metallic glasses. At the same growth temperature, the asymmetry in growth rate for rods that are heated versus cooled decreases with nanorod diameter and vanishes for very small rods. We hypothesize that structural ordering enhances crystal growth, in contrast to assumptions from common growth models. The asymmetric growth rate is attributed to the difference in the degree of the structural ordering, which is pronounced in the heated glass but sparse in the cooled liquid.

We present the structure of an engineered protein-protein interface between two beta barrel proteins, which is mediated by interactions between threonine (Thr) residues. This Thr zipper structure suggests that the protein interface is stabilized by close-packing of the Thr residues, with only one intermonomer hydrogen bond (H-bond) between two of the Thr residues. This Thr-rich interface provides a unique opportunity to study the behavior of Thr in the context of many other Thr residues. In previous work, we have shown that the side chain (χ 1 ) dihedral angles of interface and core Thr residues can be predicted with high accuracy using a hard sphere plus stereochemical constraint (HS) model. Here, we demonstrate that in the Thr-rich local environment of the Thr zipper structure, we are able to predict the χ 1 dihedral angles of most of the Thr residues. Some, however, are not well predicted by the HS model. We therefore employed explicitly solvated molecular dynamics (MD) simulations to further investigate the side chain conformations of these residues. The MD simulations illustrate the role that transient H-bonding to water, in combination with steric constraints, plays in determining the behavior of these Thr side chains. Broader Audience Statement: Protein-protein interactions are critical to life and the search for ways to disrupt adverse protein-protein interactions involved in disease is an ongoing area of drug discovery. We must better understand protein-protein interfaces, both to be able to disrupt existing ones and to engineer new ones for a variety of biotechnological applications. We have discovered and characterized an artificial Thr-rich protein-protein interface. This novel interface demonstrates a heretofore unknown property of Thr-rich surfaces: mediating protein-protein interactions., (© 2018 The Protein Society.)

The fracture toughness of glassy materials remains poorly understood. In large part, this is due to the disordered, intrinsically non-equilibrium nature of the glass structure, which challenges its theoretical description and experimental determination. We show that the notch fracture toughness of metallic glasses exhibits an abrupt toughening transition as a function of a well-controlled fictive temperature (T f ), which characterizes the average glass structure. The ordinary temperature, which has been previously associated with a ductile-to-brittle transition, is shown to play a secondary role. The observed transition is interpreted to result from a competition between the T f -dependent plastic relaxation rate and an applied strain rate. Consequently, a similar toughening transition as a function of strain rate is predicted and demonstrated experimentally. The observed mechanical toughening transition bears strong similarities to the ordinary glass transition and explains the previously reported large scatter in fracture toughness data and ductile-to-brittle transitions.

Protein core repacking is a standard test of protein modeling software. A recent study of six different modeling software packages showed that they are more successful at predicting side chain conformations of core compared to surface residues. All the modeling software tested have multicomponent energy functions, typically including contributions from solvation, electrostatics, hydrogen bonding and Lennard-Jones interactions in addition to statistical terms based on observed protein structures. We investigated to what extent a simplified energy function that includes only stereochemical constraints and repulsive hard-sphere interactions can correctly repack protein cores. For single residue and collective repacking, the hard-sphere model accurately recapitulates the observed side chain conformations for Ile, Leu, Phe, Thr, Trp, Tyr and Val. This result shows that there are no alternative, sterically allowed side chain conformations of core residues. Analysis of the same set of protein cores using the Rosetta software suite revealed that the hard-sphere model and Rosetta perform equally well on Ile, Leu, Phe, Thr and Val; the hard-sphere model performs better on Trp and Tyr and Rosetta performs better on Ser. We conclude that the high prediction accuracy in protein cores obtained by protein modeling software and our simplified hard-sphere approach reflects the high density of protein cores and dominance of steric repulsion., (© The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.)

We investigate the role of steric interactions in defining side-chain conformations in protein cores. Previously, we explored the strengths and limitations of hard-sphere dipeptide models in defining sterically allowed side-chain conformations and recapitulating key features of the side-chain dihedral angle distributions observed in high-resolution protein structures. Here, we show that modeling residues in the context of a particular protein environment, with both intra- and inter-residue steric interactions, is sufficient to specify which of the allowed side-chain conformations is adopted. This model predicts 97% of the side-chain conformations of Leu, Ile, Val, Phe, Tyr, Trp and Thr core residues to within 20°. Although the hard-sphere dipeptide model predicts the observed side-chain dihedral angle distributions for both Thr and Ser, the model including the protein environment predicts side-chain conformations to within 20° for only 60% of core Ser residues. Thus, this approach can identify the amino acids for which hard-sphere interactions alone are sufficient and those for which additional interactions are necessary to accurately predict side-chain conformations in protein cores. We also show that our approach can predict alternate side-chain conformations of core residues, which are supported by the observed electron density., (© The Author 2016. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.)

Clinical diagnosis of acute infectious diseases during the early stages of infection is critical to administering the appropriate treatment to improve the disease outcome. We present a data driven analysis of the human cellular response to respiratory viruses including influenza, respiratory syncytia virus, and human rhinovirus, and compared this with the response to the bacterial endotoxin, Lipopolysaccharides (LPS). Using an anomaly detection framework we identified pathways that clearly distinguish between asymptomatic and symptomatic patients infected with the four different respiratory viruses and that accurately diagnosed patients exposed to a bacterial infection. Connectivity pathway analysis comparing the viral and bacterial diagnostic signatures identified host cellular pathways that were unique to patients exposed to LPS endotoxin indicating this type of analysis could be used to identify host biomarkers that can differentiate clinical etiologies of acute infection. We applied the Multivariate State Estimation Technique (MSET) on two human influenza (H1N1 and H3N2) gene expression data sets to define host networks perturbed in the asymptomatic phase of infection. Our analysis identified pathways in the respiratory virus diagnostic signature as prognostic biomarkers that triggered prior to clinical presentation of acute symptoms. These early warning pathways correctly predicted that almost half of the subjects would become symptomatic in less than forty hours post-infection and that three of the 18 subjects would become symptomatic after only 8 hours. These results provide a proof-of-concept for utility of anomaly detection algorithms to classify host pathway signatures that can identify presymptomatic signatures of acute diseases and differentiate between etiologies of infection. On a global scale, acute respiratory infections cause a significant proportion of human co-morbidities and account for 4.25 million deaths annually. The development of clinical diagnostic tools to distinguish between acute viral and bacterial respiratory infections is critical to improve patient care and limit the overuse of antibiotics in the medical community. The identification of prognostic respiratory virus biomarkers provides an early warning system that is capable of predicting which subjects will become symptomatic to expand our medical diagnostic capabilities and treatment options for acute infectious diseases. The host response to acute infection may be viewed as a deterministic signaling network responsible for maintaining the health of the host organism. We identify pathway signatures that reflect the very earliest perturbations in the host response to acute infection. These pathways provide a monitor the health state of the host using anomaly detection to quantify and predict health outcomes to pathogens.

Clinicians need to predict patient outcomes with high accuracy as early as possible after disease inception. In this manuscript, we show that patient-to-patient variability sets a fundamental limit on outcome prediction accuracy for a general class of mathematical models for the immune response to infection. However, accuracy can be increased at the expense of delayed prognosis. We investigate several systems of ordinary differential equations (ODEs) that model the host immune response to a pathogen load. Advantages of systems of ODEs for investigating the immune response to infection include the ability to collect data on large numbers of 'virtual patients', each with a given set of model parameters, and obtain many time points during the course of the infection. We implement patient-to-patient variability v in the ODE models by randomly selecting the model parameters from distributions with coefficients of variation v that are centered on physiological values. We use logistic regression with one-versus-all classification to predict the discrete steady-state outcomes of the system. We find that the prediction algorithm achieves near 100% accuracy for v = 0, and the accuracy decreases with increasing v for all ODE models studied. The fact that multiple steady-state outcomes can be obtained for a given initial condition, i.e. the basins of attraction overlap in the space of initial conditions, limits the prediction accuracy for v > 0. Increasing the elapsed time of the variables used to train and test the classifier, increases the prediction accuracy, while adding explicit external noise to the ODE models decreases the prediction accuracy. Our results quantify the competition between early prognosis and high prediction accuracy that is frequently encountered by clinicians.

A fundamental question in protein science is what is the intrinsic propensity for an amino acid to be in an α-helix, β-sheet, or other backbone dihedral angle ( ϕ-ψ) conformation. This question has been hotly debated for many years because including all protein crystal structures from the protein database, increases the probabilities for α-helical structures, while experiments on small peptides observe that β-sheet-like conformations predominate. We perform molecular dynamics (MD) simulations of a hard-sphere model for Ala dipeptide mimetics that includes steric interactions between nonbonded atoms and bond length and angle constraints with the goal of evaluating the role of steric interactions in determining protein backbone conformational preferences. We find four key results. For the hard-sphere MD simulations, we show that (1) β-sheet structures are roughly three and half times more probable than α-helical structures, (2) transitions between α-helix and β-sheet structures only occur when the backbone bond angle τ (NCα C) is greater than 110°, and (3) the probability distribution of τ for Ala conformations in the "bridge" region of ϕ-ψ space is shifted to larger angles compared to other regions. In contrast, (4) the distributions obtained from Amber and CHARMM MD simulations in the bridge regions are broader and have increased τ compared to those for hard sphere simulations and from high-resolution protein crystal structures. Our results emphasize the importance of hard-sphere interactions and local stereochemical constraints that yield strong correlations between ϕ-ψ conformations and τ., (© 2014 The Protein Society.)

The physical nature of the bacterial cytoplasm is poorly understood even though it determines cytoplasmic dynamics and hence cellular physiology and behavior. Through single-particle tracking of protein filaments, plasmids, storage granules, and foreign particles of different sizes, we find that the bacterial cytoplasm displays properties that are characteristic of glass-forming liquids and changes from liquid-like to solid-like in a component size-dependent fashion. As a result, the motion of cytoplasmic components becomes disproportionally constrained with increasing size. Remarkably, cellular metabolism fluidizes the cytoplasm, allowing larger components to escape their local environment and explore larger regions of the cytoplasm. Consequently, cytoplasmic fluidity and dynamics dramatically change as cells shift between metabolically active and dormant states in response to fluctuating environments. Our findings provide insight into bacterial dormancy and have broad implications to our understanding of bacterial physiology, as the glassy behavior of the cytoplasm impacts all intracellular processes involving large components., (Copyright © 2014 Elsevier Inc. All rights reserved.)

We address the identification of optimal biomarkers for the rapid diagnosis of neonatal sepsis. We employ both canonical correlation analysis (CCA) and sparse support vector machine (SSVM) classifiers to select the best subset of biomarkers from a large hematological data set collected from infants with suspected sepsis from Yale-New Haven Hospital's Neonatal Intensive Care Unit (NICU). CCA is used to select sets of biomarkers of increasing size that are most highly correlated with infection. The effectiveness of these biomarkers is then validated by constructing a sparse support vector machine diagnostic classifier. We find that the following set of five biomarkers capture the essential diagnostic information (in order of importance): Bands, Platelets, neutrophil CD64, White Blood Cells, and Segs. Further, the diagnostic performance of the optimal set of biomarkers is significantly higher than that of isolated individual biomarkers. These results suggest an enhanced sepsis scoring system for neonatal sepsis that includes these five biomarkers. We demonstrate the robustness of our analysis by comparing CCA with the Forward Selection method and SSVM with LASSO Logistic Regression.

Background: We introduce Iterative Feature Removal (IFR) as an unbiased approach for selecting features with diagnostic capacity from large data sets. The algorithm is based on recently developed tools in machine learning that are driven by sparse feature selection goals. When applied to genomic data, our method is designed to identify genes that can provide deeper insight into complex interactions while remaining directly connected to diagnostic utility. We contrast this approach with the search for a minimal best set of discriminative genes, which can provide only an incomplete picture of the biological complexity., Results: Microarray data sets typically contain far more features (genes) than samples. For this type of data, we demonstrate that there are many equivalently-predictive subsets of genes. We iteratively train a classifier using features identified via a sparse support vector machine. At each iteration, we remove all the features that were previously selected. We found that we could iterate many times before a sustained drop in accuracy occurs, with each iteration removing approximately 30 genes from consideration. The classification accuracy on test data remains essentially flat even as hundreds of top-genes are removed.Our method identifies sets of genes that are highly predictive, even when comprised of genes that individually are not. Through automated and manual analysis of the selected genes, we demonstrate that the selected features expose relevant pathways that other approaches would have missed., Conclusions: Our results challenge the paradigm of using feature selection techniques to design parsimonious classifiers from microarray and similar high-dimensional, small-sample-size data sets. The fact that there are many subsets of genes that work equally well to classify the data provides a strong counter-result to the notion that there is a small number of "top genes" that should be used to build classifiers. In our results, the best classifiers were formed using genes with limited univariate power, thus illustrating that deeper mining of features using multivariate techniques is important.

To successfully design new proteins and understand the effects of mutations in natural proteins, we must understand the geometric and physicochemical principles underlying protein structure. The side chains of amino acids in peptides and proteins adopt specific dihedral angle combinations; however, we still do not have a fundamental quantitative understanding of why some side-chain dihedral angle combinations are highly populated and others are not. Here we employ a hard-sphere plus stereochemical constraint model of dipeptide mimetics to enumerate the side-chain dihedral angles of leucine (Leu) and isoleucine (Ile), and identify those conformations that are sterically allowed versus those that are not as a function of the backbone dihedral angles ϕ and ψ. We compare our results with the observed distributions of side-chain dihedral angles in proteins of known structure. With the hard-sphere plus stereochemical constraint model, we obtain agreement between the model predictions and the observed side-chain dihedral angle distributions for Leu and Ile. These results quantify the extent to which local, geometrical constraints determine protein side-chain conformations., (Copyright © 2013 The Authors. Published by Elsevier Inc. All rights reserved.)

Intrinsically disordered proteins (IDPs) are increasingly recognized for their important roles in a range of biological contexts, both in normal physiological function and in a variety of devastating human diseases. However, their structural characterization by traditional biophysical methods, for the purposes of understanding their function and dysfunction, has proved challenging. Here, we investigate the model IDPs α-Synuclein (αS) and tau, that are involved in major neurodegenerative conditions including Parkinson's and Alzheimer's diseases, using excluded volume Monte Carlo simulations constrained by pairwise distance distributions from single-molecule fluorescence measurements. Using this, to our knowledge, novel approach we find that a relatively small number of intermolecular distance constraints are sufficient to accurately determine the dimensions and polymer conformational statistics of αS and tau in solution. Moreover, this method can detect local changes in αS and tau conformations that correlate with enhanced aggregation. Constrained Monte Carlo simulations produce ensembles that are in excellent agreement both with experimental measurements on αS and tau and with all-atom, explicit solvent molecular dynamics simulations of αS, with much lower configurational sampling requirements and computational expense., (Copyright © 2012 Biophysical Society. Published by Elsevier Inc. All rights reserved.)

The energy functions used to predict protein structures typically include both molecular-mechanics and knowledge-based terms. In contrast, our approach is to develop robust physics- and geometry-based methods. Here, we investigate to what extent simple hard-sphere models can be used to predict side-chain conformations. The distributions of the side-chain dihedral angle χ(1) of Val and Thr in proteins of known structure show distinctive features: Val side chains predominantly adopt χ(1) = 180°, whereas Thr side chains typically adopt χ(1) = 60° and 300° (i.e., χ(1) = ±60° or g- and g(+) configurations). Several hypotheses have been proposed to explain these differences, including interresidue steric clashes and hydrogen-bonding interactions. In contrast, we show that the observed side-chain dihedral angle distributions for both Val and Thr can be explained using only local steric interactions in a dipeptide mimetic. Our results emphasize the power of simple physical approaches and their importance for future advances in protein engineering and design., (Copyright © 2012 Biophysical Society. Published by Elsevier Inc. All rights reserved.)

The pioneering work of Ramachandran and colleagues emphasized the dominance of steric constraints in specifying the structure of polypeptides. The ubiquitous Ramachandran plot of backbone dihedral angles (ϕ and ψ) defined the allowed regions of conformational space. These predictions were subsequently confirmed in proteins of known structure. Ramachandran and colleagues also investigated the influence of the backbone angle τ on the distribution of allowed ϕ/ψ combinations. The "bridge region" (ϕ ≤ 0° and -20° ≤ ψ ≤ 40°) was predicted to be particularly sensitive to the value of τ. Here we present an analysis of the distribution of ϕ/ψ angles in 850 non-homologous proteins whose structures are known to a resolution of 1.7 Å or less and sidechain B-factor less than 30 Ų. We show that the distribution of ϕ/ψ angles for all 87,000 residues in these proteins shows the same dependence on τ as predicted by Ramachandran and colleagues. Our results are important because they make clear that steric constraints alone are sufficient to explain the backbone dihedral angle distributions observed in proteins. Contrary to recent suggestions, no additional energetic contributions, such as hydrogen bonding, need be invoked., (Copyright © 2011 The Protein Society.)

We have investigated wavelength-dependent light scattering in biomimetic structures with short-range order. Coherent backscattering experiments are performed to measure the transport mean free path over a wide wavelength range. Overall scattering strength is reduced significantly due to short-range order and near-field effects. Our analysis explains why single scattering of light is dominant over multiple scattering in similar biological structures and is responsible for color generation.

A theoretical framework is developed to study the dynamics of protein folding. The key insight is that the search for the native protein conformation is influenced by the rate r at which external parameters, such as temperature, chemical denaturant, or pH, are adjusted to induce folding. A theory based on this insight predicts that 1), proteins with complex energy landscapes can fold reliably to their native state; 2), reliable folding can occur as an equilibrium or out-of-equilibrium process; and 3), reliable folding only occurs when the rate r is below a limiting value, which can be calculated from measurements of the free energy. We test these predictions against numerical simulations of model proteins with a single energy scale.