This article describes an approach to incorporate expert opinion on observable quantities through the use of a loss function which updates a prior belief as opposed to specifying parameters on the priors. Eliciting information on observable quantities allows experts to provide meaningful information on a quantity familiar to them, in contrast to elicitation on model parameters, which may be subject to interactions with other parameters or non-linear transformations before obtaining an observable quantity. The approach to incorporating expert opinion described in this paper is distinctive in that we do not specify a prior to match an expert's opinion on observed quantity, rather we obtain a posterior by updating the model parameters through a loss function. This loss function contains the observable quantity, expressed a function of the parameters, and is related to the expert's opinion which is typically operationalized as a statistical distribution. Parameters which generate observable quantities which are further from the expert's opinion incur a higher loss, allowing for the model parameters to be estimated based on their fidelity to both the data and expert opinion, with the re
The field of opinion dynamics has its roots in early research that applied methods from magnetic physics to gain insights into the formation of social opinions. A central challenge in this field lies in modeling how diverse opinions coexist and exert influence on each other. In the realm of social issues, it's In this study, we leverage the dimer construct and the dimer model to establish a theoretical framework. Through numerical simulations, we demonstrate how this proposed model can be applied to real-world scenarios of social opinion formation. The model involves the computation of the Castellain matrix (K), the distribution function (Z), and the probability of dimer configuration (P(D)) for convex regions with varying positions and distances. It explores how alterations in convex regions impact the probability of dimer configuration. Furthermore, our model takes into account two critical factors: "dependence" and "forgetting" in the process of opinion formation. It also delves into the concepts of "distance" and "location" of opinions. The results of numerical simulations shed light on how our model effectively captures the processes involved in real-world social opinion forma
Deterministic dynamics is a mathematical model used to describe the temporal evolution of a system, generally expressed as dx/dt = F(x), where x represents the system's state, and F(x) determines its dynamics. It is employed to understand long-term system behavior, including opinion formation and polarization in online communities. Opinion dynamics models, like the Katz model and the logistic map, help analyze how individual opinions are influenced within social networks and exhibit chaotic behavior. These models are crucial for studying opinion formation and collective behavior on social media, especially in conjunction with branching theory. For instance, Galam's Ising model applies principles from physics to social sciences, representing individual opinions as "spins" and illustrating how local interactions influence consensus formation. The Bounding Confidence model considers opinions within a confidence interval, showing how opinions converge or polarize. These models effectively analyze opinion dynamics in online communities, aiding in understanding trends and viral phenomena on social media. This research aims to analyze discourse flow and opinion evolution, predicting futur
Drug addiction is a complex and pervasive global challenge that continues to pose significant public health concerns. Traditional approaches to anti-addiction drug discovery have struggled to deliver effective therapeutics, facing high attrition rates, long development timelines, and inefficiencies in processing large-scale data. Artificial intelligence (AI) has emerged as a transformative solution to address these issues. Using advanced algorithms, AI is revolutionizing drug discovery by enhancing the speed and precision of key processes. This review explores the transformative role of AI in the pipeline for anti-addiction drug discovery, including data collection, target identification, and compound optimization. By highlighting the potential of AI to overcome traditional barriers, this review systematically examines how AI addresses critical gaps in anti-addiction research, emphasizing its potential to revolutionize drug discovery and development, overcome challenges, and advance more effective therapeutic strategies.
We consider the Merton problem of optimizing expected power utility of terminal wealth in the case of an unobservable Markov-modulated drift. What makes the model special is that the agent is allowed to purchase costly expert opinions of varying quality on the current state of the drift, leading to a mixed stochastic control problem with regular and impulse controls involving random consequences. Using ideas from filtering theory, we first embed the original problem with unobservable drift into a full information problem on a larger state space. The value function of the full information problem is characterized as the unique viscosity solution of the dynamic programming PDE. This characterization is achieved by a new variant of the stochastic Perron's method, which additionally allows us to show that, in between purchases of expert opinions, the problem reduces to an exit time control problem which is known to admit an optimal feedback control. Under the assumption of sufficient regularity of this feedback map, we are able to construct optimal trading and expert opinion strategies.
This study introduces a new numerical model to simulate how information is comprehended and processed on social networks, using continuous "Phase Field Modeling" variables (phiA, phiB, phiC) to represent individual users' opinions. It captures the immediate and two-way nature of social media interactions, reproducing the spread and feedback of information. The model incorporates psychological and social factors like confirmation bias and opinion rigidity to analyze information processing and opinion development among users. It also explores the dynamics of opinion segregation and interaction in and out of filter bubbles, offering a quantitative view of opinion dynamics on platforms like social networking services (SNS). This approach combines theoretical models with real-world social network data to study the effects of information concentration on opinion formation and the phenome Phase Field Modeling of opinion polarization and echo chamber effects on SNS. This paper is partially an attempt to utilize "Generative AI" and was written with educational intent. There are currently no plans for it to become a peer-reviewed paper.
Large language models (LLMs) are in the ascendancy for research in drug discovery, offering unprecedented opportunities to reshape drug research by accelerating hypothesis generation, optimizing candidate prioritization, and enabling more scalable and cost-effective drug discovery pipelines. However there is currently a lack of objective assessments of LLM performance to ascertain their advantages and limitations over traditional drug discovery platforms. To tackle this emergent problem, we have developed DrugPlayGround, a framework to evaluate and benchmark LLM performance for generating meaningful text-based descriptions of physiochemical drug characteristics, drug synergism, drug-protein interactions, and the physiological response to perturbations introduced by drug molecules. Moreover, DrugPlayGround is designed to work with domain experts to provide detailed explanations for justifying the predictions of LLMs, thereby testing LLMs for chemical and biological reasoning capabilities to push their greater use at the frontier of drug discovery at all of its stages.
Forensic experts use specialized training and knowledge to enable other members of the judicial system to make better informed and more just decisions. Factfinders, in particular, are tasked with judging how much weight to give to experts' reports and opinions. Many references describe assessing evidential weight from the perspective of a forensic expert. Some recognize that stakeholders are each responsible for evaluating their own weight of evidence. Morris (1971, 1974, 1977) provided a general framework for recipients to update their own uncertainties after learning an expert's opinion. Although this framework is normative under Bayesian axioms and several forensic scholars advocate the use of Bayesian reasoning, few resources describe its application in forensic science. This paper addresses this gap by examining how recipients can combine principles of science and Bayesian reasoning to evaluate their own likelihood ratios for expert opinions. This exercise helps clarify how an expert's role depends on whether one envisions recipients to be logical and scientific or deferential. Illustrative examples with an expert's opinion expressed as a categorical conclusion, likelihood rat
In this highly digitised world, fake news is a challenging problem that can cause serious harm to society. Considering how fast fake news can spread, automated methods, tools and services for assisting users to do fact-checking (i.e., fake news detection) become necessary and helpful, for both professionals, such as journalists and researchers, and the general public such as news readers. Experts, especially researchers, play an essential role in informing people about truth and facts, which makes them a good proxy for non-experts to detect fake news by checking relevant expert opinions and comments. Therefore, in this paper, we present aedFaCT, a web browser extension that can help professionals and news readers perform fact-checking via the automatic discovery of expert opinions relevant to the news of concern via shared keywords. Our initial evaluation with three independent testers (who did not participate in the development of the extension) indicated that aedFaCT can provide a faster experience to its users compared with traditional fact-checking practices based on manual online searches, without degrading the quality of retrieved evidence for fact-checking. The source code o
While deep learning has revolutionized computer-aided drug discovery, the AI community has predominantly focused on model innovation and placed less emphasis on establishing best benchmarking practices. We posit that without a sound model evaluation framework, the AI community's efforts cannot reach their full potential, thereby slowing the progress and transfer of innovation into real-world drug discovery. Thus, in this paper, we seek to establish a new gold standard for small molecule drug discovery benchmarking, WelQrate. Specifically, our contributions are threefold: WelQrate Dataset Collection - we introduce a meticulously curated collection of 9 datasets spanning 5 therapeutic target classes. Our hierarchical curation pipelines, designed by drug discovery experts, go beyond the primary high-throughput screen by leveraging additional confirmatory and counter screens along with rigorous domain-driven preprocessing, such as Pan-Assay Interference Compounds (PAINS) filtering, to ensure the high-quality data in the datasets; WelQrate Evaluation Framework - we propose a standardized model evaluation framework considering high-quality datasets, featurization, 3D conformation generat
Introduction: Artificial intelligence (AI) is exhibiting tremendous potential to reduce the massive costs and long timescales of drug discovery. There are however important challenges currently limiting the impact and scope of AI models. Areas covered: In this perspective, the authors discuss a range of data issues (bias, inconsistency, skewness, irrelevance, small size, high dimensionality), how they challenge AI models, and which issue-specific mitigations have been effective. Next, they point out the challenges faced by uncertainty quantification techniques aimed at enhancing and trusting the predictions from these AI models. They also discuss how conceptual errors, unrealistic benchmarks and performance misestimation can confound the evaluation of models and thus their development. Lastly, the authors explain how human bias, whether from AI experts or drug discovery experts, constitutes another challenge that can be alleviated by gaining more prospective experience. Expert opinion: AI models are often developed to excel on retrospective benchmarks unlikely to anticipate their prospective performance. As a result, only a few of these models are ever reported to have prospective
The role of Artificial Intelligence (AI) is growing in every stage of drug development. Nevertheless, a major challenge in drug discovery AI remains: Drug pharmacokinetic (PK) and Drug-Target Interaction (DTI) datasets collected in different studies often exhibit limited overlap, creating data overlap sparsity. Thus, data curation becomes difficult, negatively impacting downstream research investigations in high-throughput screening, polypharmacy, and drug combination. We propose xImagand-DKI, a novel SMILES/Protein-to-Pharmacokinetic/DTI (SP2PKDTI) diffusion model capable of generating an array of PK and DTI target properties conditioned on SMILES and protein inputs that exhibit data overlap sparsity. We infuse additional molecular and genomic domain knowledge from the Gene Ontology (GO) and molecular fingerprints to further improve our model performance. We show that xImagand-DKI-generated synthetic PK data closely resemble real data univariate and bivariate distributions, and can adequately fill in gaps among PK and DTI datasets. As such, xImagand-DKI is a promising solution for data overlap sparsity and may improve performance for downstream drug discovery research tasks. Code
In a social network, individuals express their opinions on several interdependent topics, and therefore the evolution of their opinions on these topics is also mutually dependent. In this work, we propose a differential game model for the multi-dimensional opinion formation of a social network whose population of agents interacts according to a communication graph. Each individual's opinion evolves according to an aggregation of disagreements between the agent's opinions and its graph neighbors on multiple interdependent topics exposed to an unknown extraneous disturbance. For a social network with strategist agents the opinions evolve over time with respect to the minimization of a quadratic cost function that solely represents each individual's motives against the disturbance. We find the unique Nash/worst-case equilibrium solution for the proposed differential game model of coupled multi-dimensional opinions under an open-loop information structure. Moreover, we propose a distributed implementation of the Nash/worst-case equilibrium solution. We examine the non-distributed and proposed distributed open-loop Nash/worst-case strategies on a small social network with strategist age
Despite decades of advancements in automated ligand screening, large-scale drug discovery remains resource-intensive and requires post-processing hit selection, a step where chemists manually select a few promising molecules based on their chemical intuition. This creates a major bottleneck in the virtual screening process for drug discovery, demanding experts to repeatedly balance complex trade-offs among drug properties across a vast pool of candidates. To improve the efficiency and reliability of this process, we propose a novel human-centered framework named CheapVS that allows chemists to guide the ligand selection process by providing preferences regarding the trade-offs between drug properties via pairwise comparison. Our framework combines preferential multi-objective Bayesian optimization with a docking model for measuring binding affinity to capture human chemical intuition for improving hit identification. Specifically, on a library of 100K chemical candidates targeting EGFR and DRD2, CheapVS outperforms state-of-the-art screening methods in identifying drugs within a limited computational budget. Notably, our method can recover up to 16/37 EGFR and 37/58 DRD2 known drug
The proliferation of public networks has enabled instantaneous and interactive communication that transcends temporal and spatial constraints. The vast amount of textual data on the Web has facilitated the study of quantitative analysis of public opinion, which could not be visualized before. In this paper, we propose a new theory of opinion dynamics. This theory is designed to explain consensus building and opinion splitting in opinion exchanges on social media such as Twitter. With the spread of public networks, immediate and interactive communication that transcends temporal and spatial constraints has become possible, and research is underway to quantitatively analyze the distribution of public opinion, which has not been visualized until now, using vast amounts of text data. In this paper, we propose a model based on the Like Bounded Confidence Model, which represents opinions as continuous quantities. However, the Bounded Confidence mModel assumes that people with different opinions move without regard to their opinions, rather than ignoring them. Furthermore, our theory modeled the phenomenon in such a way that it can incorporate and represent the effects of external externa
The field of opinion dynamics has evolved steadily since the earliest studies applying magnetic physics methods to better understand social opinion formation. However, in the real world, complete agreement of opinions is rare, and biaxial consensus, especially on social issues, is rare. To address this challenge, Ishii and Kawabata (2018) proposed an extended version of the Bounded Confidence Model that introduces new parameters indicating dissent and distrust, as well as the influence of mass media. Their model aimed to capture more realistic social opinion dynamics by introducing coefficients representing the degree of trust and distrust, rather than assuming convergence of opinions. In this paper, we propose a new approach to opinion dynamics based on this Trust-Distrust Model (TDM), applying the dimer allocation and Ising model. Our goal is to explore how the interaction between trust and distrust affects social opinion formation. In particular, we analyze through mathematical models how various external stimuli, such as mass media, third-party opinions, and economic and political factors, affect people's opinions. Our approach is to mathematically represent the dynamics of tru
The simulations in this paper are based on the theory of opinion dynamics, which incorporates both Opinion A and Opinion B, a case that is the inverse of Opinion A, in human relationships. It was confirmed that aspects of consensus building depend on the ratio of the trust coefficient to the distrust coefficient. In this study, the ratio of trust to distrust tended to vary like a phase transition around 55%, but we wanted to see if the same phenomenon could be confirmed in large-scale cases. In the previous case studies, this tendency has been observed from N = 300 to N = 1600 , and we will discuss the case of N = 10000 with N = 3000. By verifying the extent of the phenomenon on the social scale, we intend to consider simulation items for consensus building, such as consideration of the sensitivities of topics in online and offline opinion formation.
Accurate drug target affinity prediction can improve drug candidate selection, accelerate the drug discovery process, and reduce drug production costs. Previous work focused on traditional fingerprints or used features extracted based on the amino acid sequence in the protein, ignoring its 3D structure which affects its binding affinity. In this work, we propose GraphPrint: a framework for incorporating 3D protein structure features for drug target affinity prediction. We generate graph representations for protein 3D structures using amino acid residue location coordinates and combine them with drug graph representation and traditional features to jointly learn drug target affinity. Our model achieves a mean square error of 0.1378 and a concordance index of 0.8929 on the KIBA dataset and improves over using traditional protein features alone. Our ablation study shows that the 3D protein structure-based features provide information complementary to traditional features.
In the expansive realm of drug discovery, with approximately 15,000 known drugs and only around 4,200 approved, the combinatorial nature of the chemical space presents a formidable challenge. While Artificial Intelligence (AI) has emerged as a powerful ally, traditional AI frameworks face significant hurdles. This manuscript introduces CardiGraphormer, a groundbreaking approach that synergizes self-supervised learning (SSL), Graph Neural Networks (GNNs), and Cardinality Preserving Attention to revolutionize drug discovery. CardiGraphormer, a novel combination of Graphormer and Cardinality Preserving Attention, leverages SSL to learn potent molecular representations and employs GNNs to extract molecular fingerprints, enhancing predictive performance and interpretability while reducing computation time. It excels in handling complex data like molecular structures and performs tasks associated with nodes, pairs of nodes, subgraphs, or entire graph structures. CardiGraphormer's potential applications in drug discovery and drug interactions are vast, from identifying new drug targets to predicting drug-to-drug interactions and enabling novel drug discovery. This innovative approach prov
In decision modelling with time to event data, there are a variety of parametric models which could be used to extrapolate the survivor function. Each of these implies a different hazard function and in situations where there is moderate censoring, they can result in quite different extrapolations. Expert opinion on the long-term survival or other quantities could reduce model uncertainty. We present a general and easily implementable approach for including a variety of types of expert opinions. Expert opinion is incorporated by penalizing the likelihood function. Inference is performed in a Bayesian framework, however, this approach can also be implemented using frequentist methods. The issue of aggregating pooling expert opinions is also considered and included in the analysis. We validate the method against a previously published approach and include a worked example of this method. This work highlights that expert opinions can be implemented in a straightforward manner using this approach, however, more work is required on the correct elicitation of these quantities.