Biased Sampling Methods Research Articles

Taking the Next Step in Exploring the Literary Digest 1936 Poll

While many instructors are aware of the Literary Digest 1936 poll as an example of biased sampling methods, this article details potential further explorations for the Digest’s 1924–1936 quadrennial U.S. presidential election polls. Potential activities range from lessons in data acquisition, cleaning, and validation, to basic data literacy and visualization skills, to exploring one or more methods of adjustment to account for bias based on information collected at that time. Students can also compare how those methods would have performed. One option could be to give introductory students a first look at the idea of “sampling adjustment” and how this principle can be used to account for difficulties in modern polling, but the context is rich in other opportunities that can be discussed at various times in the course or in more advanced sampling courses. Supplementary materials for this article are available online.

Adaptive Sampling Methods for Molecular Dynamics in the Era of Machine Learning.

Molecular dynamics (MD) simulations are fundamental computational tools for the study of proteins and their free energy landscapes. However, sampling protein conformational changes through MD simulations is challenging due to the relatively long time scales of these processes. Many enhanced sampling approaches have emerged to tackle this problem, including biased sampling and path-sampling methods. In this Perspective, we focus on adaptive sampling algorithms. These techniques differ from other approaches because the thermodynamic ensemble is preserved and the sampling is enhanced solely by restarting MD trajectories at particularly chosen seeds rather than introducing biasing forces. We begin our treatment with an overview of theoretically transparent methods, where we discuss principles and guidelines for adaptive sampling. Then, we present a brief summary of select methods that have been applied to realistic systems in the past. Finally, we discuss recent advances in adaptive sampling methodology powered by deep learning techniques, as well as their shortcomings.

High-throughput virtual screening of potential inhibitors of GPR52 using docking and biased sampling method for Huntington's disease therapy.

Huntington's disease (HD) is a rare and progressive neurodegenerative disorder caused by polyglutamine (poly-Q) mutations of the huntingtin (HTT) gene resulting in chorea, cognitive, and psychiatric dysfunctions. Being a monogenic condition, reducing the levels of the mutated huntingtin protein (mHTT) holds promise as an effective therapeutic approach. GPR52, an orphan G-protein coupled receptor (GPCR), enriched in the striatum, is a novel target for slowing down the progression of HD by lowering the mHTT levels. Therefore, the study focuses on identifying potent small-molecule inhibitors for GPR52 using a combination of robust high-throughput virtual screening (HTVS) and pharmacokinetics profiling followed by fast pulling of ligand (FPL) and umbrella sampling (US) simulations. Initially, screening a library of 2,36,545 compounds was done against the binding pocket of GPR52. Based on binding affinity, stereochemical and non-bonded interactions, and pharmacokinetic profiling, 50 compounds were shortlisted. Selected hit compounds 1, 2, and 3 were subjected to FPL simulations with applied external bias potential to investigate their unique dissociation pathways and intermolecular interactions over time. Subsequently, the US simulations were performed on the selected hit compounds to estimate their binding free energy (ΔG). The analysis of the trajectories obtained from simulations revealed that the residues TYR34, TYR185, GLY187, ASP188, ILE189, SER299, PHE300, and THR303 within the active site of GPR52 were significant for efficient ligand binding through the formation of various hydrogen bond interactions and hydrophobic contacts. Out of the three hit compounds, compound 3 had the lowest ΔG of - 20.82 ± 0.44kcal/mol. The study identified compounds 1, 2, and 3 as potential molecules that can be developed as GPR52 inhibitors holding promise for lowering mHTT levels.

An Efficient Robot Exploration Method Based on Heuristics Biased Sampling

Autonomous robot exploration has received widespread attention as a crucial step in constructing maps for unknown environments. Among the research results in the field of robotic exploration, the boundary exploration algorithm based on Rapidly-exploring Random Tree (RRT) performs well in most scenarios, which guides the robot motion by taking the frontier with the maximum revenue value as the target point discretely. However, this discrete target selection scheme ignores the geometric continuity of the environment, and the robot often turns to another area before completing a geometrically continuous area (e.g. a complete room), which results the disorderly motion of the robot and a relatively low exploration efficiency. This work proposes a heuristics biased sampling based robot exploration strategy, which utilizes the semantic information of the environment as the heuristics to guide the robot exploration. Firstly, a lightweight network model is proposed for heuristic object recognition. Secondly, a geometrically continuous region is estimated based on the heuristic object, and frontiers are then extracted in this region by using a biased sampling method. Finally, the heuristic information gain model is designed to determine the target frontier for the exploration, which instructs the robot to select the frontiers in the heuristic region with a higher priority. In this way, the robot can effectively use the heuristic knowledge of the environment to improve the efficiency of exploration. Compared with the RRT based exploration methods in simulations and real-world environmental studies, and the experimental results prove the feasibility and effectiveness of our approach.

In-silico insights into aminoglycoside intake by TMC1.

Aminoglycosides are small-molecule antibiotics that are widely used as treatment against a variety of bacterial infections. A side-effect of aminoglycosides use is dose-dependent hearing loss, often referred to as cochleotoxicity. This is hypothesized to be due to the inability of the inner-ear sensory hair cells to clear these molecules out leading to permanent hair-cell death. Multiple experiments involving manipulations of components of the hair-cell mechanotransduction machinery, including tip links and transmembrane channel-like 1 (TMC1) and TMC2 proteins, have shown that an important pathway for aminoglycoside intake by hair-cells is their transduction channel. Recent studies have also shown that TMC1 and TMC2 proteins are the pore forming components of the hair-cell mechanotransduction machinery. Although experiments suggest that these positively charged aminoglycoside molecules can enter hair cells through TMCs, the mechanism and the energetics of this process have not been studied. Here we use AlphaFold2 models of the human TMC1 protein with and without calcium- and integrin-binding protein 2 (CIB2) in the presence of Kanamycin A (KanA) along with microsecond-long voltage simulations to study the pore opening mechanism and the interactions between KanA and pore residues. We also present the free-energy surface of the process obtained using metadynamics simulations (a biased sampling method) which quantifies the energetics of this pore-crossing event. The open-pore conformations that were generated because of the KanA crossing event were further simulated with voltage to calculate the conductance of the channel in its putative open state and to assess the role of CIB2 in its conduction. These results give qualitative and a quantitative insights into the mechanism of pore opening and KanA crossing through the inner-ear transduction channel.

확률 비용 맵을 이용한 수확 로봇의 경로 계획 알고리즘

Due to the increasing global demand for food and shortage of agricultural labor, researchers have focused on developing new technologies, capable of increasing production yields and farming efficiency. Since the repetitive harvesting task is labor-intensive, a robotic harvesting system is one of the promising solutions. To enable precise robotic harvesting, it is important to determine a path for the robot that avoids damaging the crops and other obstacles such as unripe crops, leaves, and stems. In this paper, we present a probabilistic costmap classified according to obstacle type and use it in an RRT-star path planning algorithm. The proposed path planning algorithm has two main features: a biased sampling method based on a probabilistic costmap and a costmap-based collision detector. We have validated the proposed path planning method in the scenario of cucumber harvesting tasks using a two-dimensional planar robot.

Protein Structure Validation Derives a Smart Conformational Search in a Physically Relevant Configurational Subspace.

Since proteins perform biological functions through their dynamic properties, molecular dynamics (MD) simulation is a sophisticated strategy for investigating their functions. Analyses of trajectories provide statistical information about a specific protein as a free-energy landscape (FEL). However, the timescale of normal MD is shorter than that of biological functions, resulting in statistically insufficient conformational sampling, finally leading to unreliable FEL calculation. To search for a broad configurational subspace, an external bias is imposed on a target protein as biased sampling. However, its regulation is challenging because the optimal strength of the perturbation is unknown. Furthermore, a physically irrelevant configurational subspace was searched when imposing an inappropriate external bias. To address this issue, we newly proposed an external biased regulation scheme known as the G-factor external bias limiter (GERBIL). In GERBIL, protein configurations generated by external bias are structurally validated by an indicator (G-factor), enabling the search for a physically relevant subspace. In addition to biased sampling, nonbiased sampling might search for a physically irrelevant configurational subspace because repeating multiple MD simulations from several initial structures tends to search for an overly broad configurational subspace. For this issue, the structural qualities of configurations generated by nonbiased sampling have not been investigated. Therefore, we confirmed whether the G-factor screened the collapsed (low-quality) configurations generated by nonbiased sampling. To address this issue, the outlier flooding method (OFLOOD) was adopted in GERBIL as a nonbiased sampling method, which is referred to as OFLOOD-GERBIL. OFLOOD rapidly expands a configurational subspace by resampling the rarely occurring states of a given protein and tends to search an overly broad subspace. Thus, we considered that GERBIL might improve the excessive conformational search of OFLOOD for a physically irrelevant configurational subspace. As a demonstration, OFLOOD and OFLOOD-GERBIL were applied to a globular protein (T4 lysozyme) and their conformational search qualities were assessed. Based on our assessment, normal OFLOOD without the outlier validation frequently sampled low-quality configurations, whereas OFLOOD-GERBIL with the outlier validation intensively sampled high-quality configurations. In conclusion, OFLOOD-GERBIL derives a smart conformational search in a physically relevant configurational subspace, indicating that protein structure validation works in both nonbiased and biased sampling methods.

Molecular Driving Force for Facet Selectivity of Sequence-Defined Amphiphilic Peptoids at Au-Water Interfaces.

Shape-controlled colloidal nanocrystal syntheses often require facet-selective solution-phase chemical additives to regulate surface free energy, atom addition/migration fluxes, or particle attachment rates. Because of their highly tunable properties and robustness to a wide range of experimental conditions, peptoids represent a very promising class of next-generation functional additives for control over nanocrystal growth. However, understanding the origin of facet selectivity at the molecular level is critical to generalizing their design. Herein we employ molecular dynamics simulations and biased sampling methods and report stronger selectivity to Au(111) than to Au(100) for Nce3Ncp6, a peptoid that has been shown to assist the formation of 5-fold twinned Au nanostars. We find that facet selectivity is achieved through synergistic effects of both peptoid-surface and solvent-surface interactions. Moreover, the amphiphilic nature of Nce3Ncp6 together with the order of peptoid-peptoid and peptoid-surface binding energies, that is, peptoid-Au(100) < peptoid-peptoid < peptoid-Au(111), further amplifies its distinct collective behavior on different Au surfaces. Our studies provide a fundamental understanding of the molecular origin of facet-selective adsorption and highlight the possibility of future designs and uses of sequence-defined peptoids for predictive syntheses of nanocrystals with designed shapes and properties.

Calculation Methods of Solution Chemical Potential and Application in Emulsion Microencapsulation.

Several new biased sampling methods were summarized for solution chemical potential calculation methods in the field of emulsion microencapsulation. The principles, features, and calculation efficiencies of various biased Widom insertion sampling methods were introduced, including volume detection bias, simulation ensemble bias, and particle insertion bias. The proper matches between various types of solution in emulsion and biased Widom methods were suggested, following detailed analyses on the biased insertion techniques. The volume detection bias methods effectively improved the accuracy of the data and the calculation efficiency by inserting detection particles and were suggested to be used for the calculation of solvent chemical potential for the homogeneous aqueous phase of the emulsion. The chemical potential of water, argon, and fluorobenzene (a typical solvent of the oil phase in double emulsion) was calculated by a new, optimized volume detection bias proposed by this work. The recently developed Well-Tempered(WT)-Metadynamics method skillfully constructed low-density regions for particle insertion and dynamically adjusted the system configuration according to the potential energy around the detection point, and hence, could be used for the oil-polymer mixtures of microencapsulation emulsion. For the macromolecule solutes in the oil or aqueous phase of the emulsion, the particle insertion bias could be applied to greatly increase the success rate of Widom insertions. Readers were expected to choose appropriate biased Widom methods to carry out their calculations on chemical potential, fugacity, and solubility of solutions based on the system molecular properties, inspired by this paper.

The impact of biased sampling of event logs on the performance of process discovery

With Process discovery algorithms, we discover process models based on event data, captured during the execution of business processes. The process discovery algorithms tend to use the whole event data. When dealing with large event data, it is no longer feasible to use standard hardware in a limited time. A straightforward approach to overcome this problem is to down-size the data utilizing a random sampling method. However, little research has been conducted on selecting the right sample, given the available time and characteristics of event data. This paper systematically evaluates various biased sampling methods and evaluates their performance on different datasets using four different discovery techniques. Our experiments show that it is possible to considerably speed up discovery techniques using biased sampling without losing the resulting process model quality. Furthermore, due to the implicit filtering (removing outliers) obtained by applying the sampling technique, the model quality may even be improved.

The Adjustment of Covariates in Cox’s Model under Case-Cohort Design

Case-cohort design is a biased sampling method. Due to its cost-effective and theoretical significance, this design has extensive application value in many large cohort studies. The case-cohort data includes a subcohort sampled randomly from the entire cohort and all the failed subjects outside the subcohort. In this paper, the adjustment for the distorted covariates is considered to case-cohort data in Cox’s model. According to the existing adjustable methods of distorted covariates for linear and nonlinear models, we propose estimating the distorting functions by nonparametrically regressing the distorted covariates on the distorting factors; then, the estimators for the parameters are obtained using the estimated covariates. The proof of consistency and being asymptotically normal is completed. For calculating the maximum likelihood estimates of the regression coefficients subject in Cox’s model, a minorization-maximization (MM) algorithm is developed. Simulation studies are performed to compare the estimations with the covariates undistorted, distorted, and adjusted to illustrate the proposed methods.

Prediction of AChE-ligand affinity using the umbrella sampling simulation

Acetylcholinesterase (AChE) is characterized as a key target for designing inhibitors to prevent Alzheimer's disease (AD). The binding free energy of a ligand to the AChE enzyme is a critical factor to screen the potential inhibitor in addition to pharmacokinetics and pharmacology estimation. The biased sampling or umbrella sampling (US) method emerges as a reliable technique to estimate the AChE-inhibitor affinity. The affinity is computed as the difference between the largest and smallest values of the free energy change, obtained by using a potential of mean force (PMF) analysis. The obtained affinities overestimate the experimental ones with a value of ∼4.10 kcal/mol. However, a very good correlation coefficient (R=0.94) between the computational and experimental values is observed. Consequently, the obtained precision is high since the mean error of the free energy value is of δ=1.17 kcal/mol. The binding affinity of a new ligand can be consistently appraised via the US technique. Therefore, the absolute binding free energy of a ligand to the AChE protein can be obtained via the linear regression with the root-mean-square errors (RMSE) of 0.98 kcal/mol. The small value of RMSE implies that ligands revealing the similar binding affinities are able to be discriminated through the US simulations. In addition, the derivatives of Cordyceps were recently reported that they are able to inhibit the AChE enzyme, resulting in an improvement in learning and cognitive ability for curing AD. The active metabolites of Cordyceps were thus evaluated as the potential inhibitors for the AChE enzyme, and the 5-Carboxy-2′-deoxyuridine compound can inhibit the activity of the AChE enzyme. These compounds are also passed the testing of Lipinski's rule of five, toxicity, crossing blood-brain barrier (BBB) ability, and human intestinal absorption.

C-Terminal Plays as the Possible Nucleation of the Self-Aggregation of the S-Shape Aβ11-42 Tetramer in Solution: Intensive MD Study.

Amyloid beta (Aβ) peptides are characterized as the major factors associated with neuron death in Alzheimer’s disease, which is listed as the most common form of neurodegeneration. Disordered Aβ peptides are released from proteolysis of the amyloid precursor protein. The Aβ self-assembly process roughly takes place via five steps: disordered forms → oligomers → photofibrils → mature fibrils → plaques. Although Aβ fibrils are often observed in patient brains, oligomers were recently indicated to be major neurotoxic elements. In this work, the neurotoxic compound S-shape Aβ11–42 tetramer (S4Aβ11–42) was investigated over 10 μs of unbiased MD simulations. In particular, the S4Aβ11–42 oligomer adopted a high dynamics structure, resulting in unsuccessful determination of their structures in experiments. The C-terminal was suggested as the possible nucleation of the Aβ42 aggregation. The sequences 27–35 and 39–40 formed rich β-content, whereas other residues mostly adopted coil structures. The mean value of the β-content over the equilibrium interval is ∼42 ± 3%. Furthermore, the dissociation free energy of the S4Aβ11–42 peptide was predicted using a biased sampling method. The obtained free energy is ΔGUS = −58.44 kcal/mol which is roughly the same level as the corresponding value of the U-shape Aβ17-42 peptide. We anticipate that the obtained S4Aβ11–42 structures could be used as targets for AD inhibitor screening over the in silico study.

Effective Estimation of Ligand-Binding Affinity Using Biased Sampling Method.

The binding between two biomolecules is one of the most critical factors controlling many bioprocesses. Therefore, it is of great interest to derive a reliable method to calculate the free binding energy between two biomolecules. In this work, we have demonstrated that the binding affinity of ligands to proteins can be determined through biased sampling simulations. The umbrella sampling (US) method was applied on 20 protein–ligand complexes, including the cathepsin K (CTSK), type II dehydroquinase (DHQase), heat shock protein 90 (HSP90), and factor Xa (FXa) systems. The ligand-binding affinity was evaluated as the difference between the largest and smallest values of the free-energy curve, which was obtained via a potential of mean force analysis. The calculated affinities differ sizably from the previously reported experimental values, with an average difference of ∼3.14 kcal/mol. However, the calculated results are in good correlation with the experimental data, with correlation coefficients of 0.76, 0.87, 0.96, and 0.97 for CTSK, DHQase, HSP90, and FXa, respectively. Thus, the binding free energy of a new ligand can be reliably estimated using our US approach. Furthermore, the root-mean-square errors (RMSEs) of binding affinity of these systems are 1.13, 0.90, 0.37, and 0.25 kcal/mol, for CTSK, DHQase, HSP90, and FXa, respectively. The small RMSE values indicate the good precision of the biased sampling method that can distinguish the ligands exhibiting similar binding affinities.

Free Energy Landscape for the Entire Transport Cycle of Triose-Phosphate/Phosphate Translocator

Secondary active transporters translocate their substrates using the electrochemical potentials of other chemicals and undergo large-scale conformational changes. Despite extensive structural studies, the atomic details of the transport mechanism still remain elusive. We performed a series of all-atom molecular dynamics simulations of the triose-phosphate/phosphate translocator (TPT), which exports organic phosphates in the chloroplast stroma in strict counter exchange with inorganic phosphate (Pi). Biased sampling methods, including the string method and umbrella sampling, successfully reproduced the conformational changes between the inward- and outward-facing states, along with the substrate binding. The free energy landscape of this entire TPT transition pathway demonstrated the alternating access and substrate translocation mechanisms, which revealed that Pi is relayed by positively charged residues along the transition pathway. Furthermore, the conserved Glu207 functions as a "molecular switch", linking the local substrate binding and the global conformational transition. Our results provide atomic-detailed insights into the substrate transport mechanism of the antiporter.

Hydrological seasonality dictates fish fauna of the lower Araguaia River, Tocantins-Araguaia basin

Hydrological seasonality of flow dictates diversity in tropical aquatic ecosystems. Large tropical rivers and tributaries are typically more depth and shows increased flow velocity and area in flood than in dry season. A thorough analysis of seasonality effect on fish community structure can provide valuable information of major forces driving tropical communities. However, these types of analyses have been limited by poor knowledge of tropical diversity, human impacts on environment, and biased sampling methods. We used a pristine large tropical river (and tributaries) and five types (12 gill nets, beach seine nets, longlines, branch baited hooks, and cast nets) of fishing gears to present an evaluation of flow seasonality effect on fish community. We sampled fourteen sites in river channel and tributaries in the lower Araguaia River. Sampling was conducted in flood (March and May 2009) and dry (July and September 2009) seasons of a typical seasonal year. Species richness, Shannon Diversity and Evenness, mean similarity of binary data, and abundance (CPUE) were analyzed by PERMANOVA and PCoA. We also analyzed the species most correlated (by Spearman correlation rank) with flood and dry season. Features differed significantly between flood and dry seasons, but not between channel and tributaries. The composition and abundance of community, and the representative species of flood and dry changed quite along the year. Flood showed lotic-related species, and dry presented lentic-related species. Thus, we achieved an alternation of fish community between season, both for channel and tributaries.

Fresh is best: Accurate SNP genotyping from koala scats.

Maintaining genetic diversity is a crucial component in conserving threatened species. For the iconic Australian koala, there is little genetic information on wild populations that is not either skewed by biased sampling methods (e.g., sampling effort skewed toward urban areas) or of limited usefulness due to low numbers of microsatellites used. The ability to genotype DNA extracted from koala scats using next‐generation sequencing technology will not only help resolve location sample bias but also improve the accuracy and scope of genetic analyses (e.g., neutral vs. adaptive genetic diversity, inbreeding, and effective population size). Here, we present the successful SNP genotyping (1272 SNP loci) of koala DNA extracted from scat, using a proprietary DArTseq™ protocol. We compare genotype results from two‐day‐old scat DNA and 14‐day‐old scat DNA to a blood DNA template, to test accuracy of scat genotyping. We find that DNA from fresher scat results in fewer loci with missing information than DNA from older scat; however, 14‐day‐old scat can still provide useful genetic information, depending on the research question. We also find that a subset of 209 conserved loci can accurately identify individual koalas, even from older scat samples. In addition, we find that DNA sequences identified from scat samples through the DArTseq™ process can provide genetic identification of koala diet species, bacterial and viral pathogens, and parasitic organisms.

Biased sampling from facebook multilayer activity network using learning automata

Although much research has been devoted to unbiased sampling of various networks, bias is not always disadvantageous, but sometimes useful. Especially for many real-world applications such as detecting influential nodes, spam users, and the most trustful people, it is preferred to sample users with special properties. Since sampling from friendship network alone cannot collect these important nodes appropriately, one may use interactions occurred among users. This paper deals with biased sampling of multilayer activity network. The proposed method initially learns the transition probabilities according to the considered application using learning automata. Then we sample the graph by running an application-based random walk following the learnt probabilities, in order to be guided to suitable nodes and collect their information. At last, the performance of the proposed method in terms of different applications such as fame, spam, and trust is evaluated and compared with those of common sampling algorithms. According to the experiments done, biased sampling method based on learning automata outperforms all other sampling approaches including simple random walk, Metropolis-Hastings random walk, BFS, forest fire, degree, and uniform sampling in terms of all the evaluation measures. To the best of our knowledge, our method is the first and only biased sampling method which can be used in a multilayer activity network.

Stochastic Properties of the Weighted Hazard Rate Order

The hazard rate order between two weighted distributions with a given weight function is considered as a new stochastic order called weighted hazard rate order. The influences of biased sampling methods on properties of the hazard rate order are usefully pointed out with this weighted order. We first obtain some basic properties of the weighted hazard rate order and then develop some useful preservation properties of it with respect to monotone transformations, mixture of distributions, convolution of distributions, order statistics and residual lifetime of distributions. A few examples are also presented to show the usefulness of the results achieved in the paper.

Active-Site Hydration and Water Diffusion in Cytochrome P450cam: A Highly Dynamic Process

Long-timescale molecular dynamics simulations (300 ns) are performed on both the apo- (i.e., camphor-free) and camphor-bound cytochrome P450cam (CYP101). Water diffusion into and out of the protein active site is observed without biased sampling methods. During the course of the molecular dynamics simulation, an average of 6.4 water molecules is observed in the camphor-binding site of the apo form, compared to zero water molecules in the binding site of the substrate-bound form, in agreement with the number of water molecules observed in crystal structures of the same species. However, as many as 12 water molecules can be present at a given time in the camphor-binding region of the active site in the case of apo-P450cam, revealing a highly dynamic process for hydration of the protein active site, with water molecules exchanging rapidly with the bulk solvent. Water molecules are also found to exchange locations frequently inside the active site, preferentially clustering in regions surrounding the water molecules observed in the crystal structure. Potential-of-mean-force calculations identify thermodynamically favored trans-protein pathways for the diffusion of water molecules between the protein active site and the bulk solvent. Binding of camphor in the active site modifies the free-energy landscape of P450cam channels toward favoring the diffusion of water molecules out of the protein active site.