Computational Cognitive Models

Many computational methods to predict the macromolecular targets of small organic molecules have been presented to date. Despite progress, target prediction methods still have important limitations. For example, the most accurate methods implicitly restrict their predictions to a relatively small number of targets, are not systematically validated on drugs (whose targets are harder to predict than those of non-drug molecules) and often lack a reliability score associated with each predicted target. Here we present a systematic validation of ligand-centric target prediction methods on a set of clinical drugs. These methods exploit a knowledge-base covering 887,435 known ligand-target associations between 504,755 molecules and 4,167 targets. Based on this dataset, we provide a new estimate of the polypharmacology of drugs, which on average have 11.5 targets below IC50 10 µM. The average performance achieved across clinical drugs is remarkable (0.348 precision and 0.423 recall, with la...

The focus of studies on second-order false belief reasoning generally was on investigating the roles of executive functions and language with correlational studies. Different from those studies, we focus on the question how 5-year-olds select and revise reasoning strategies in second-order false belief tasks by constructing two computational cognitive models of this process: an instance-based learning model and a reinforcement learning model. Unlike the reinforcement learning model, the instance-based learning model predicted that children who fail second-order false belief tasks would give answers based on first-order theory of mind (ToM) reasoning as opposed to zero-order reasoning. This prediction was confirmed with an empirical study that we conducted with 72 5- to 6-year-old children. The results showed that 17% of the answers were correct and 83% of the answers were wrong. In line with our prediction, 65% of the wrong answers were based on a first-order ToM strategy, while onl...

Agents need to be able to adapt to changes in their environment. One way to achieve this, is to provide agents with the ability of self-modification. Self-modification requires reflection and strategies with which new knowledge can be acquired, a necessary condition for creativity. This paper describes a knowledge-level model for the design of self-modifying agents and explores the feasibility of automatically designing self-modifying agents. * To appear in Proceedings of the Creative Design Workshop, December 2001. * relate description of agent to properties of agent * relate required properties to required properties * relate required properties to structure modifications of an agent * understand tasks, the ontologies, knowledge, and control involved * make strategic decisions within the agent's own process control * determine which strategies to deploy within an agent * resolve conflicts within an agent and strategies are needed to * guide the overall design process * alternate between viewpoints on design requirements and agent descriptions * prioritize possible strategies for self-modification. The process of designing a self-modifying information retrieval agent, introduced in Section 3.3, is used as an illustration. The self-modifying information retrieval agent was designed on the basis of the following requirements: 1. The agent is able to find information on web-pages. 2. The agent is able to adapt itself. 3. The agent is able to communicate with other agents on queries and query results. These requirements state that the information retrieval agent is able to find information on the web. To design such an agent, the design process needs specific knowledge of * designing an information retrieval agent * self-modification capabilities * reflective capabilities * modification of information retrieval agents.

With the growing understanding of complex diseases, the focus of drug discovery has shifted away from the wellaccepted "one target, one drug" model, to a new "multi-target, multi-drug" model, aimed at systemically modulating multiple targets. Identification of the interaction between drugs and target proteins plays an important role in genomic drug discovery, in order to discover new drugs or novel targets for existing drugs. Due to the laborious and costly experimental process of drug-target interaction prediction, in silico prediction could be an efficient way of providing useful information in supporting experimental interaction data. An important notion that has emerged in postgenomic drug discovery is that the large-scale integration of genomic, proteomic, signaling and metabolomic data can allow us to construct complex networks of the cell that would provide us with a new framework for understanding the molecular basis of physiological or pathophysiological states. An emerging paradigm of polypharmacology in the postgenomic era is that drug, target and disease spaces can be correlated to study the effect of drugs on different spaces and their interrelationships can be exploited for designing drugs or cocktails which can effectively target one or more disease states. The future goal, therefore, is to create a computational platform that integrates genome-scale metabolic pathway, protein-protein interaction networks, gene transcriptional analysis in order to build a comprehensive network for multi-target multi-drug discovery.

We can understand and act upon the beliefs of other people, even when these conflict with our own beliefs. Children's development of this ability, known as Theory of Mind, typically happens around age 4. Research using a looking-time paradigm, however, established that toddlers at the age of 15 months old pass a non-verbal false-belief task (Onishi and Baillargeon in Science 308:255-258, 2005). This is well before the age at which children pass any of the verbal false-belief tasks. In this study we present a more complex case of false-belief reasoning with older children. We tested second-order reasoning, probing children's ability to handle the belief of one person about the belief of another person. We find just the opposite: 7-year-olds pass a verbal false-belief reasoning task, but fail on an equally complex low-verbal task. This finding suggests that language supports explicit reasoning about beliefs, perhaps by facilitating the cognitive system to keep track of beliefs attributed by people to other people.

The focus of studies on second-order false belief reasoning generally was on investigating the roles of executive functions and language with correlational studies. Different from those studies, we focus on the question how 5-year-olds select and revise reasoning strategies in second-order false belief tasks by constructing two computational cognitive models of this process: an instance-based learning model and a reinforcement learning model. Unlike the reinforcement learning model, the instance-based learning model predicted that children who fail second-order false belief tasks would give answers based on first-order theory of mind (ToM) reasoning as opposed to zero-order reasoning. This prediction was confirmed with an empirical study that we conducted with 72 5- to 6-year-old children. The results showed that 17% of the answers were correct and 83% of the answers were wrong. In line with our prediction, 65% of the wrong answers were based on a first-order ToM strategy, while onl...

With the growing understanding of complex diseases, the focus of drug discovery has shifted away from the wellaccepted "one target, one drug" model, to a new "multi-target, multi-drug" model, aimed at systemically modulating multiple targets. Identification of the interaction between drugs and target proteins plays an important role in genomic drug discovery, in order to discover new drugs or novel targets for existing drugs. Due to the laborious and costly experimental process of drug-target interaction prediction, in silico prediction could be an efficient way of providing useful information in supporting experimental interaction data. An important notion that has emerged in postgenomic drug discovery is that the large-scale integration of genomic, proteomic, signaling and metabolomic data can allow us to construct complex networks of the cell that would provide us with a new framework for understanding the molecular basis of physiological or pathophysiological states. An emerging paradigm of polypharmacology in the postgenomic era is that drug, target and disease spaces can be correlated to study the effect of drugs on different spaces and their interrelationships can be exploited for designing drugs or cocktails which can effectively target one or more disease states. The future goal, therefore, is to create a computational platform that integrates genome-scale metabolic pathway, protein-protein interaction networks, gene transcriptional analysis in order to build a comprehensive network for multi-target multi-drug discovery.

In this paper we propose a computational framework aimed at extending the problem solving capabilities of cognitive artificial agents through the introduction of a novel, goal-directed, dynamic knowledge generation mechanism obtained via a non monotonic reasoning procedure. In particular, the proposed framework relies on the assumption that certain classes of problems cannot be solved by simply learning or injecting new external knowledge in the declarative memory of a cognitive artificial agent but, on the other hand, require a mechanism for the automatic and creative re-framing, or re-formulation, of the available knowledge. We show how such mechanism can be obtained trough a framework of dynamic knowledge generation that is able to tackle the problem of commonsense concept combination. In addition, we show how such a framework can be employed in the field of cognitive architectures in order to overcome situations like the impasse in SOAR by extending the possible options of its subgoaling procedures.

The focus of studies on second-order false belief reasoning generally was on investigating the roles of executive functions and language with correlational studies. Different from those studies, we focus on the question how 5-year-olds select and revise reasoning strategies in second-order false belief tasks by constructing two computational cognitive models of this process: an instance-based learning model and a reinforcement learning model. Unlike the reinforcement learning model, the instance-based learning model predicted that children who fail second-order false belief tasks would give answers based on first-order theory of mind (ToM) reasoning as opposed to zero-order reasoning. This prediction was confirmed with an empirical study that we conducted with 72 5-to 6-year-old children. The results showed that 17% of the answers were correct and 83% of the answers were wrong. In line with our prediction, 65% of the wrong answers were based on a first-order ToM strategy, while only 29% of them were based on a zero-order strategy (the remaining 6% of subjects did not provide any answer). Based on our instance-based learning model, we propose that when children get feedback " Wrong, " they explicitly revise their strategy to a higher level instead of implicitly selecting one of the available ToM strategies. Moreover, we predict that children's failures are due to lack of experience and that with exposure to second-order false belief reasoning, children can revise their wrong first-order reasoning strategy to a correct second-order reasoning strategy.

During the last decades, many cognitive architectures (CAs) have been realized adopting different assumptions about the organization and the representation of their knowledge level. Some of them (e.g. SOAR [35]) adopt a classical symbolic approach, some (e.g. LEABRA[48]) are based on a purely con-nectionist model, while others (e.g. CLARION [59]) adopt a hybrid approach combining connectionist and symbolic representational levels. Additionally, some attempts (e.g. biSOAR) trying to extend the representational capacities of CAs by integrating diagrammatical representations and reasoning are also available [34]. In this paper we propose a reflection on the role that Conceptual Spaces, a framework developed by Peter Gärdenfors [24] more than fifteen years ago, can play in the current development of the Knowledge Level in Cognitive Systems and Architectures. In particular, we claim that Conceptual Spaces offer a lingua franca that allows to unify and generalize many aspects of the symbolic, sub-symbolic and diagrammatic approaches (by overcoming some of their typical problems) and to integrate them on a common ground. In doing so we extend and detail some of the arguments explored by Gärdenfors [23] for defending the need of a conceptual, intermediate, representation level between the symbolic and the sub-symbolic one. In particular we focus on the advantages offered by Conceptual Spaces (w.r.t. symbolic and sub-symbolic approaches) in dealing with the problem of compositionality of representations based on typicality traits. Additionally, we argue that Conceptual Spaces could offer a unifying framework for interpreting many kinds of diagrammatic and analogical representations. As a consequence, their adoption could also favor the integration of diagrammatical representation and reasoning in CAs.

In this article we present an advanced version of Dual-PECCS, a cognitively-inspired knowledge representation and reasoning system aimed at extending the capabilities of artificial systems in conceptual categorization tasks. It combines different sorts of common-sense cat-egorization (prototypical and exemplars-based categorization) with standard monotonic cate-gorization procedures. These different types of inferential procedures are reconciled according to the tenets coming from the dual process theory of reasoning. On the other hand, from a representational perspective, the system relies on the hypothesis of conceptual structures represented as heterogeneous proxytypes. Dual-PECCS has been experimentally assessed in a task of conceptual categorization where a target concept illustrated by a simple common-sense linguistic description had to be identified by resorting to a mix of categorization strategies, and its output has been compared to human responses. The obtained results suggest that our approach can be beneficial to improve the representational and reasoning conceptual capabilities of standard cognitive artificial systems, and –in addition– that it may be plausibly applied to different general computational models of cognition. The current version of the system , in fact, extends our previous work, in that Dual-PECCS is now integrated and tested into two cognitive architectures, ACT-R and CLARION, implementing different assumptions on the underlying invariant structures governing human cognition. Such integration allowed us to extend our previous evaluation.

We overview the main historical and technological elements characterising the rise, the fall and the recent renaissance of the cognitive approaches to Artificial Intelligence and provide some insights and suggestions about the future directions and challenges that, in our opinion, this discipline needs to face in the next years.

Polypharmacology has emerged as novel means in drug discovery for improving treatment response in clinical use. However, to really capitalize on the polypharmacological effects of drugs, there is a critical need to better model and understand how the complex interactions between drugs and their cellular targets contribute to drug efficacy and possible side effects. Network graphs provide a convenient modeling framework for dealing with the fact that most drugs act on cellular systems through targeting multiple proteins both through on-target and off-target binding. Network pharmacology models aim at addressing questions such as how and where in the disease network should one target to inhibit disease phenotypes, such as cancer growth, ideally leading to therapies that are less vulnerable to drug resistance and side effects by means of attacking the disease network at the systems level through synergistic and synthetic lethal interactions. Since the exponentially increasing number of potential drug target combinations makes pure experimental approach quickly unfeasible, this review depicts a number of computational models and algorithms that can effectively reduce the search space for determining the most promising combinations for experimental evaluation. Such computational-experimental strategies are geared toward realizing the full potential of multi-target treatments in different disease phenotypes. Our specific focus is on system-level network approaches to polypharmacology designs in anticancer drug discovery, where we give representative examples of how network-centric modeling may offer systematic strategies toward better understanding and even predicting the phenotypic responses to multi-target therapies.

Explanations in cognitive science rely predominantly on
computational modeling. Though the scientific practice is
systematic, and there is little doubt about the empirical value
of numerous models, the methodological account of
computational explanation is not up-to-date. The current
paper offers a systematic account of computational
explanation in cognitive science in a largely mechanistic
framework. The account is illustrated with a short case study
of modeling of the mirror neuron system in terms of
predictive coding.

Recent accounts of mechanistic explanation suggest that models are only explanatory insofar as they map neatly onto lower level components (Craver, 2006; Piccinini and Craver, 2011; Kaplan and Craver, 2011). In the case of extensions into the cognitive sciences, this has meant interpreting the ‘integration’ of cognitive psychology and neuroscience to mean treating cognitive models as scaffolding for more detailed neural mechanisms. Bechtel (2008) and Piccinini and Craver (2011) describe this sort of integration as a seamless process where the functionally-defined components of cognitive models are localized in parts of the brain. One critic of this view (Weiskopf, 2011) has responded by denying that cognitive models provide mechanistic explanations, and arguing for the independent value of functional explanations of cognitive phenomena.
I think the problem with these accounts of integration lies elsewhere. Models may be mechanistic even if they don’t align tidily with lower-level ones. The problem with seamless accounts of cognitive-neural integration is their seamlessness, not that they take cognitive models to be mechanistic. The non-componential view of mechanisms described in Machamer et al (2000) allows for cognitive and neural models that cross cut one another, and for cognitive models that don’t necessarily decompose into neural parts.
I critique three aspects of seamless accounts of integration: the claim that cognitive models are elliptical mechanism sketches; the assumption that cognitive mechanisms must be decomposable and localizable in neural parts; and the heuristic of decomposition and localization being promoted to the status of a normative constraint on mechanisms. The integrations we can realistically expect are more partial, and patchy than the mosaic unity Craver (2007) describes.

In this paper we develop a conceptual modelling framework for knowledge-level reflection (KLR), i.e, the modelling of tasks that require a self-representation of a knowledge system's own object-level problem solving tasks. This framework builds upon the KADS methodology for knowledge acquisition and design of knowledge systems .

We can understand and act upon the beliefs of other people, even when these conflict with our own beliefs. Children's development of this ability, known as Theory of Mind, typically happens around age 4. Research using a looking-time paradigm, however, established that toddlers at the age of 15 months old pass a non-verbal false-belief task (Onishi and Baillargeon in Science 308: [255][256][257][258] 2005). This is well before the age at which children pass any of the verbal false-belief tasks. In this study we present a more complex case of false-belief reasoning with older children. We tested second-order reasoning, probing children's ability to handle the belief of one person about the belief of another person. We find just the opposite: 7-year-olds pass a verbal false-belief reasoning task, but fail on an equally complex low-verbal task. This finding suggests that language supports explicit reasoning about beliefs, perhaps by facilitating the cognitive system to keep track of beliefs attributed by people to other people.