Small Ensembles Research Articles

Predicting Affinity Through Homology (PATH): Interpretable Binding Affinity Prediction with Persistent Homology.

Accurate binding affinity prediction is crucial to structure-based drug design. Recent work used computational topology to obtain an effective representation of protein-ligand interactions. While algorithms using algebraic topology have proven useful in predicting properties of biomolecules, previous algorithms employed uninterpretable machine learning models which failed to explain the underlying geometric and topological features that drive accurate binding affinity prediction. Moreover, they had high computational complexity which made them intractable for large proteins. We present the fastest known algorithm to compute persistent homology features for protein-ligand complexes using opposition distance, with a runtime that is independent of the protein size. Then, we exploit these features in a novel, interpretable algorithm to predict protein-ligand binding affinity. Our algorithm achieves interpretability through an effective embedding of distances across bipartite matchings of the protein and ligand atoms into real-valued functions by summing Gaussians centered at features constructed by persistent homology. We name these functions internuclear persistent contours (IPCs) . Next, we introduce persistence fingerprints , a vector with 10 components that sketches the distances of different bipartite matching between protein and ligand atoms, refined from IPCs. Let the number of protein atoms in the protein-ligand complex be n , number of ligand atoms be m , and ω ≈ 2.4 be the matrix multiplication exponent. We show that for any 0 < ε < 1, after an 𝒪 ( mn log( mn )) preprocessing procedure, we can compute an ε -accurate approximation to the persistence fingerprint in 𝒪 ( m log 6 ω ( m/ε )) time, independent of protein size. This is an improvement in time complexity by a factor of 𝒪 (( m + n ) 3 ) over any previous binding affinity prediction that uses persistent homology. We show that the representational power of persistence fingerprint generalizes to protein-ligand binding datasets beyond the training dataset. Then, we introduce PATH , Predicting Affinity Through Homology, a two-part algorithm consisting of PATH + and PATH - . PATH + is an interpretable, small ensemble of shallow regression trees for binding affinity prediction from persistence fingerprints. We show that despite using 1,400-fold fewer features, PATH + has comparable performance to a previous state-of-the-art binding affinity prediction algorithm that uses persistent homology. Moreover, PATH + has the advantage of being interpretable. We visualize the features captured by persistence fingerprint for variant HIV-1 protease complexes and show that persistence fingerprint captures binding-relevant structural mutations. PATH - , in turn, uses regression trees over IPCs to differentiate between binding and decoy complexes. Finally, we benchmarked PATH versus established binding affinity prediction algorithms spanning physics-based, knowledge-based, and deep learning methods, revealing that PATH has comparable or better performance with less overfitting, compared to these state-of-the-art methods. The source code for PATH is released open-source as part of the osprey protein design software package.

Hybrid covariance super-resolution data assimilation

AbstractThe super-resolution data assimilation (SRDA) enhances a low-resolution (LR) model with a Neural Network (NN) that has learned the differences between high and low-resolution models offline and performs data assimilation in high-resolution (HR). The method enhances the accuracy of the EnKF-LR system for a minor computational overhead. However, performance quickly saturates when the ensemble size is increased due to the error introduced by the NN. We therefore combine the SRDA with the mixed-resolution data assimilation method (MRDA) into a method called “Hybrid covariance super-resolution data assimilation” (Hybrid SRDA). The forecast step runs an ensemble at two resolutions (high and low). The assimilation is done in the HR space by performing super-resolution on the LR members with the NN. The assimilation uses the hybrid covariance that combines the emulated and dynamical HR members. The scheme is extensively tested with a quasi-geostrophic model in twin experiments, with the LR grid being twice coarser than the HR. The Hybrid SRDA outperforms the SRDA, the MRDA, and the EnKF-HR at a given computational cost. The benefit is the largest compared to the EnKF-HR for small ensembles. However, even with larger computational resources, using a mix of high and low-resolution members is worth it. Besides, the Hybrid SRDA, the EnKF-HR, and the SRDA, unlike the MRDA, prevent the smoothing of dynamical structures of the background error covariance matrix. The Hybrid SRDA method is also attractive because it is customizable to available resources.

Rapid rebalancing of co-tuned ensemble activity in the auditory cortex.

Sensory information is represented by small varying neuronal ensembles in sensory cortices. In the auditory cortex (AC) repeated presentations of the same sound activate differing ensembles indicating high trial-by trial variability in activity even though the sounds activate the same percept. Efficient processing of complex acoustic signals requires that these sparsely distributed neuronal ensembles actively interact in order to provide a constant percept. Thus, the differing ensembles might interact to process the incoming sound inputs. Here, we probe interactions within and across ensembles by combining in vivo 2-photon Ca2+ imaging and holographic optogenetic stimulation to study how increased activity of single cells level affects the cortical network. We stimulated a small number of neurons sharing the same frequency preference alongside the presentation of a target pure tone, further increasing their tone-evoked activity. We found that other non-stimulated co-tuned neurons decreased their tone-evoked activity when the frequency of the presented pure tone matched to their tuning property, while non co-tuned neurons were unaffected. Activity decrease was greater for non-stimulated co-tuned neurons with higher frequency selectivity. Co-tuned and non co-tuned neurons were spatially intermingled. Our results shows that co-tuned ensembles communicated and balanced their total activity across the larger network. The rebalanced network activity due to external stimulation remained constant. These effects suggest that co-tuned ensembles in AC interact and rapidly rebalance their activity to maintain encoding homeostasis, and that the rebalanced network is persistent.

A robust, fiber-coupled scanning probe magnetometer using electron spins at the tip of a diamond nanobeam

Abstract Fiber-coupled sensors are well suited for sensing and microscopy in hard-to-reach environments such as biological or cryogenic systems. We demonstrate fiber-based magnetic imaging based on nitrogen-vacancy (NV) sensor spins at the tip of a fiber-coupled diamond nanobeam. We incorporated angled ion implantation into the nanobeam fabrication process to realize a small ensemble of NV spins at the nanobeam tip. By gluing the nanobeam to a tapered fiber, we created a robust and transportable probe with optimized optical coupling efficiency. We demonstrate the imaging capability of the fiber-coupled nanobeam by measuring the magnetic field generated by a current-carrying wire. With its robust coupling and efficient readout at the fiber-coupled interface, our probe could allow new studies of (quantum) materials and biological samples.

Efficient entanglement purification based on noise guessing decoding

In this paper, we propose a novel bipartite entanglement purification protocol built upon hashing and upon the guessing random additive noise decoding (GRAND) approach recently devised for classical error correction codes. Our protocol offers substantial advantages over existing hashing protocols, requiring fewer qubits for purification, achieving higher fidelities, and delivering better yields with reduced computational costs. We provide numerical and semi-analytical results to corroborate our findings and provide a detailed comparison with the hashing protocol of Bennet et al. Although that pioneering work devised performance bounds, it did not offer an explicit construction for implementation. The present work fills that gap, offering both an explicit and more efficient purification method. We demonstrate that our protocol is capable of purifying states with noise on the order of 10% per Bell pair even with a small ensemble of 16 pairs. The work explores a measurement-based implementation of the protocol to address practical setups with noise. This work opens the path to practical and efficient entanglement purification using hashing-based methods with feasible computational costs. Compared to the original hashing protocol, the proposed method can achieve some desired fidelity with a number of initial resources up to one hundred times smaller. Therefore, the proposed method seems well-fit for future quantum networks with a limited number of resources and entails a relatively low computational overhead.

Using Satellite Data Assimilation Techniques to Combine Infrasound Observations and a Full Ray-Tracing Model to Constrain Stratospheric Variables

Abstract Infrasound waves generated at Earth’s surface can reach high altitudes before returning to the surface to be recorded by microbarometer array stations. These waves carry information about the propagation medium, in particular temperature and winds in the atmosphere. It is only recently that studies on the assimilation of such data into atmospheric models have been published. Intending to advance this line of research, we here use the modulated ensemble transform Kalman filter (METKF)—commonly used in satellite data assimilation—to assimilate infrasound-related observations in order to update a column of three vertically varying variables: temperature and horizontal wind speeds. This includes stratospheric and mesospheric heights, which are otherwise poorly observed. The numerical experiments on synthetic data but with realistic reanalysis product atmospheric specifications (following the observing system simulation experiment paradigm) reveal that a large ensemble is capable of reducing errors, especially for wind speeds in stratospheric heights close to 30–60 km. While using a small ensemble leads to incorrect analysis increments and large estimation errors, the METKF ameliorates this problem and even achieves error reduction from the prior to the posterior mean estimator. Significance Statement The stratosphere and mesosphere have significantly less observational coverage compared to the troposphere, especially for the winds. This lack of information can reduce the accuracy of medium-range weather forecasts. By mimicking a realistic setup, this study paves the way for including novel observations in the estimation of the atmospheric state in these heights using an ensemble data assimilation method. These observations come from a dataset of opportunity containing infrasound-related measurements that are routinely carried out at several stations around the world.

Autonomous 4? Learners’ Collaborative Problem Solving in Chamber Music Rehearsals With and Without a Coach

This investigation explored collaborative problem solving with and without a coach by analyzing problem-solving (PS) and problem-solving-prompting (PSP) behaviors by group members and coaches in six total coached rehearsals with three collegiate string quartets, and six autonomous rehearsals with the same string quartets. Teachers’ and learners’ PS and PSP behaviors were compared in three settings: autonomous chamber music rehearsals, chamber music rehearsals with a coach, and one-to-one instruction. Coaches and students performed the same problem-solving component behaviors and most of the same PSP behaviors as observed in one-to-one instruction and in rehearsals of a professional chamber ensemble. As in one-to-one instruction, coaches could engage learners in problem solving, and in this setting invite them to collaborate with one another, by asking questions and varying the specificity of feedback and directives. Coaches’ Attention Directives preceded the most learner problem solving and collaborative engagement. During autonomous rehearsals learners often switched leadership roles fluidly, often prompting one another to problem solve. Implications for a pedagogy of coaching and inclusion of small ensembles within music education are discussed.

Studying phonon coherence with a quantum sensor

Nanomechanical oscillators offer numerous advantages for quantum technologies. Their integration with superconducting qubits shows promise for hardware-efficient quantum error-correction protocols involving superpositions of mechanical coherent states. Limitations of this approach include mechanical decoherence processes, particularly two-level system (TLS) defects, which have been widely studied using classical fields and detectors. In this manuscript, we use a superconducting qubit as a quantum sensor to perform phonon number-resolved measurements on a piezoelectrically coupled phononic crystal cavity. This enables a high-resolution study of mechanical dissipation and dephasing in coherent states of variable size (n¯≃1−10\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\bar{n}\\simeq 1-10$$\\end{document} phonons). We observe nonexponential relaxation and state size-dependent reduction of the dephasing rate, which we attribute to TLS. Using a numerical model, we reproduce the dissipation signatures (and to a lesser extent, the dephasing signatures) via emission into a small ensemble (N = 5) of rapidly dephasing TLS. Our findings comprise a detailed examination of TLS-induced phonon decoherence in the quantum regime.

Implications of size dispersion on X-ray scattering of crystalline nanoparticles: CeO2 as a case study

Controlling the shape and size dispersivity and crystallinity of nanoparticles (NPs) has been a challenge in identifying these parameters' role in the physical and chemical properties of NPs. The need for reliable quantitative tools for analyzing the dispersivity and crystallinity of NPs is a considerable problem in optimizing scalable synthesis routes capable of controlling NP properties. The most common tools are electron microscopy (EM) and X-ray scattering techniques. However, each technique has different susceptibility to these parameters, implying that more than one technique is necessary to characterize NP systems with maximum reliability. Wide-angle X-ray scattering (WAXS) is mandatory to access information on crystallinity. In contrast, EM or small-angle X-ray scattering (SAXS) is required to access information on whole NP sizes. EM provides average values on relatively small ensembles in contrast to the bulk values accessed by X-ray techniques. Besides the fact that the SAXS and WAXS techniques have different susceptibilities to size distributions, SAXS is easily affected by NP–NP interaction distances. Because of all the variables involved, there have yet to be proposed methodologies for cross-analyzing data from two techniques that can provide reliable quantitative results of dispersivity and crystallinity. In this work, a SAXS/WAXS-based methodology is proposed for simultaneously quantifying size distribution and degree of crystallinity of NPs. The most reliable easy-to-access size result for each technique is demonstrated by computer simulation. Strategies on how to compare these results and how to identify NP–NP interaction effects underneath the SAXS intensity curve are presented. Experimental results are shown for cubic-like CeO2 NPs. WAXS size results from two analytical procedures are compared, line-profile fitting of individual diffraction peaks in opposition to whole pattern fitting. The impact of shape dispersivity is also evaluated. Extension of the proposed methodology for cross-analyzing EM and WAXS data is possible.

Establishment and validation of an artificial intelligence-based model for real-time detection and classification of colorectal adenoma

Colorectal cancer (CRC) prevention requires early detection and removal of adenomas. We aimed to develop a computational model for real-time detection and classification of colorectal adenoma. Computationally constrained background based on real-time detection, we propose an improved adaptive lightweight ensemble model for real-time detection and classification of adenomas and other polyps. Firstly, we devised an adaptive lightweight network modification and effective training strategy to diminish the computational requirements for real-time detection. Secondly, by integrating the adaptive lightweight YOLOv4 with the single shot multibox detector network, we established the adaptive small object detection ensemble (ASODE) model, which enhances the precision of detecting target polyps without significantly increasing the model's memory footprint. We conducted simulated training using clinical colonoscopy images and videos to validate the method's performance, extracting features from 1148 polyps and employing a confidence threshold of 0.5 to filter out low-confidence sample predictions. Finally, compared to state-of-the-art models, our ASODE model demonstrated superior performance. In the test set, the sensitivity of images and videos reached 87.96% and 92.31%, respectively. Additionally, the ASODE model achieved an accuracy of 92.70% for adenoma detection with a false positive rate of 8.18%. Training results indicate the effectiveness of our method in classifying small polyps. Our model exhibits remarkable performance in real-time detection of colorectal adenomas, serving as a reliable tool for assisting endoscopists.

Design of earth-abundant Ni3ZnC0.7@Ni@C catalyst for selective butadiene hydrogenation

The pursuit of developing catalysts from earth-abundant materials to supplant those based on precious metals is of paramount importance in selective hydrogenations. While nickel-based systems have shown promise in the selective hydrogenation of butadiene, their practical applications are hampered by severe deactivation issues due to coke deposition and excessive hydrogenation. Here, a novel catalyst, Ni3ZnC0.7@Ni@C, is ingeniously engineered through the controlled oxidation of Ni3ZnC0.7@C. This catalyst is characterized by small Ni0 ensembles elegantly embellishing the Ni3ZnC0.7 nanoparticles, all encased within porous carbon shells. The evolutions of this catalyst, in terms of composition and structure during the oxidation process, is meticulously observed and characterized using a spectrum of advanced techniques. The Ni3ZnC0.7@Ni@C catalyst exhibits outstanding activity and stability in the hydrogenation of butadiene, surpassing other Ni-based systems, including its precursor Ni3ZnC0.7@C and other previously documented catalysts such as Ni3InC0.5 and the Ni3In alloy. A pivotal finding of this research is the self-limiting behavior of coke deposition in the initial reaction stages. This intriguing phenomenon not only curbs further deactivation but also significantly enhances butene production, maintaining operational stability for an impressive duration of 80 hours. This discovery underscores the advantageous role of in situ generated ‘soft’ cokes in augmenting the selectivity and stability of the catalyst, which is particularly enlightening for other catalytic processes that are similarly afflicted by coking issues, thereby opening avenues for further in-depth investigations in this field.

Ce-promoted Ni-NiO small ensemble constrained in an MgO catalyst for efficient hydrogen production through NH3 decomposition

Ce-promoted Ni-NiO small ensemble constrained in an MgO catalyst for efficient hydrogen production through NH3 decomposition

Identification of predictive patient characteristics for assessing the probability of COVID-19 in-hospital mortality.

As the world emerges from the COVID-19 pandemic, there is an urgent need to understand patient factors that may be used to predict the occurrence of severe cases and patient mortality. Approximately 20% of SARS-CoV-2 infections lead to acute respiratory distress syndrome caused by the harmful actions of inflammatory mediators. Patients with severe COVID-19 are often afflicted with neurologic symptoms, and individuals with pre-existing neurodegenerative disease have an increased risk of severe COVID-19. Although collectively, these observations point to a bidirectional relationship between severe COVID-19 and neurologic disorders, little is known about the underlying mechanisms. Here, we analyzed the electronic health records of 471 patients with severe COVID-19 to identify clinical characteristics most predictive of mortality. Feature discovery was conducted by training a regularized logistic regression classifier that serves as a machine-learning model with an embedded feature selection capability. SHAP analysis using the trained classifier revealed that a small ensemble of readily observable clinical features, including characteristics associated with cognitive impairment, could predict in-hospital mortality with an accuracy greater than 0.85 (expressed as the area under the ROC curve of the classifier). These findings have important implications for the prioritization of clinical measures used to identify patients with COVID-19 (and, potentially, other forms of acute respiratory distress syndrome) having an elevated risk of death.

All-Electrical Readout of Coherently Controlled Spins in Silicon Carbide.

Spin defects in silicon carbide are promising candidates for quantum sensing applications as they exhibit long coherence times even at room temperature. However, spin readout methods that rely on fluorescence detection can be challenging due to poor photon collection efficiency. Here, we demonstrate coherent spin control and all-electrical readout of a small ensemble of spins in a SiC junction diode using pulsed electrically detected magnetic resonance. A lock-in detection scheme based on a three stage modulation cycle is implemented, significantly enhancing the signal-to-noise ratio. This technique enabled observation of coherent spin dynamics, specifically Rabi spin nutation, spin dephasing, and spin decoherence. The use of these protocols for magnetometry applications is evaluated.

Listening to the structure: Mass-timber construction at the Groton Hill Music Center

Historically, Western music spaces were built with massive masonry-bearing walls. While these designs offer structural and acoustical benefits, today’s construction culture is conscious of more fluid forms and embodied carbon, which influences how consultants think about the design of spaces for music. The recently completed Groton Hill Music Center is unique for its use of mass timber’s sculptural capabilities to shape performance space acoustic volumes forthrightly, with little additional finish treatment. The center features studio classrooms for students of all ages, an orchestral rehearsal space, a 300-seat recital hall for soloists and small ensembles, and a grand 1000-seat concert hall that opens to view for a 500-seat lawn audience. Structure and architecture are largely unified, with one’s acoustic experience defined by the sculpted mass timber structure. The comparatively light-weight timber/concrete composite structure represents a significant and somewhat acoustically risky reduction in mass to provide an acoustically warm environment. This presentation will explore how the structure was optimized to provide stiffness and multi-scale diffusion, taking the ideals of classic concert hall design, and embracing contemporary construction technologies and acoustical analyses.

Sensing of magnetic field effects in radical-pair reactions using a quantum sensor

Magnetic field effects (MFE) in certain chemical reactions have been well established in the last five decades and are attributed to the evolution of transient radical pairs whose spin dynamics are determined by local and external magnetic fields. The majority of existing experimental techniques used to probe these reactions only provide ensemble averaged reaction parameters and spin chemistry, hindering the observation of the potential presence of quantum coherent phenomena at the single molecule scale. Here, considering a single nitrogen-vacancy (NV) center as quantum sensor, we investigate the prospects and requirements for the detection of MFEs on the spin dynamics of radical pairs at the scale of a single and small ensembles of molecules. We employ elaborate and realistic models of radical pairs, considering their coupling to the local spin environment and the sensor. For two model systems, we derive signals of MFE detectable even in the weak coupling regime between a radical pair and an NV quantum sensor and observe that the dynamics of certain populations, as well as coherence elements, of the density matrix of the radical pair are directly detectable. Our investigations will provide important guidelines for the potential detection of spin chemistry of biomolecules at the scale of a single and small ensembles of molecules. Published by the American Physical Society 2024

Non-radiative configurations of a few quantum emitters ensembles: Evolutionary optimization approach

Suppressing the spontaneous emission in quantum emitters ensembles (atoms) is one of the topical problems in quantum optics and quantum technology. While many approaches are based on utilizing the subradiance effect in ordered quantum emitters arrays, the ensemble configurations providing the minimal spontaneous emission rate are yet unknown. In this work, we employ the differential evolution algorithm to identify the optimal configurations of a few atomic ensembles that support quantum states with maximal radiative lifetime. We demonstrate that atoms tend to assemble mostly in quasi-regular structures with specific geometry, which strongly depends on the minimally allowed interatomic distance rmin. While the discovered specific non-radiative realizations of small ensembles cannot be immediately predicted, there is particular correspondence to the non-radiative states in the atomic lattices. In particular, we have found that states inheriting their properties either from the bound states in the continuum or band edge states of infinite lattices dominate across a wide range of rmin values. Additionally, we show that for small interatomic distances, the linear arrays with modulated spacing have the smallest radiative losses exponentially decreasing as the size of the ensemble increases.

Hierarchical Pruning of Deep Ensembles with Focal Diversity

Deep neural network ensembles combine the wisdom of multiple deep neural networks to improve the generalizability and robustness over individual networks. It has gained increasing popularity to study and apply deep ensemble techniques in the deep learning community. Some mission-critical applications utilize a large number of deep neural networks to form deep ensembles to achieve desired accuracy and resilience, which introduces high time and space costs for ensemble execution. However, it still remains a critical challenge whether a small subset of the entire deep ensemble can achieve the same or better generalizability and how to effectively identify these small deep ensembles for improving the space and time efficiency of ensemble execution. This article presents a novel deep ensemble pruning approach, which can efficiently identify smaller deep ensembles and provide higher ensemble accuracy than the entire deep ensemble of a large number of member networks. Our hierarchical ensemble pruning approach (HQ) leverages three novel ensemble pruning techniques. First, we show that the focal ensemble diversity metrics can accurately capture the complementary capacity of the member networks of an ensemble team, which can guide ensemble pruning. Second, we design a focal ensemble diversity based hierarchical pruning approach, which will iteratively find high quality deep ensembles with low cost and high accuracy. Third, we develop a focal diversity consensus method to integrate multiple focal diversity metrics to refine ensemble pruning results, where smaller deep ensembles can be effectively identified to offer high accuracy, high robustness and high ensemble execution efficiency. Evaluated using popular benchmark datasets, we demonstrate that the proposed hierarchical ensemble pruning approach can effectively identify high quality deep ensembles with better classification generalizability while being more time and space efficient in ensemble decision making. We have released the source codes on GitHub at https://github.com/git-disl/HQ-Ensemble .

Luján Pérez y un calvario atribuido al imaginero: la Virgen, Jesucristo y san Juan, testigos de la estela del autor en la historia del arte en Canarias.

This article approaches again the figure of the imager Luján Pérez, analyzing one of his minor sculptures attributed to him: an ordeal hitherto unknown. It is a small sculptural ensemble that concentrates the skill of the carver in a few centimeters. With the revision of the work of the guiense Luján, we emphasize one of the artists who has had more transcendence in the History of Art in the Canary Island and, as we will demonstrate here, an author of which pieces continue to appear beyond the distant dates in which the gubia joined the wood to perpetuate a superb set of religious images belonging forever to the Canarian heritage.

Chaotic Diffusion in a Triaxial Galactic Model: an Example of Global Stable Chaos

In this work we focus on the chaotic diffusion in the phase space of a triaxial potential resembling an elliptical galaxy. The transport process is studied in two different action-like starting spaces in order to cope with circulating and noncirculating orbits. Estimates of the diffusion rate obtained by means of the variance approach are discussed in detail and their limitations are exposed. After revisiting the Shannon-entropy-based method from a conceptual point of view in the framework of simple arguments taken from the information theory, we apply it to measure changes in the unperturbed actions or integrals of motion of the system for different sets of small ensembles of random initial conditions. For such sets of ensembles, estimates of the Lyapunov times are also provided. The results show that, within the chaotic component of the phase space, the Lyapunov times are shorter than any physical time scale as the Hubble time, but the diffusion times are much larger than the latter. Thus, we conclude that stable chaos dominates the dynamics of realistic galactic models.