Attention Selectively Reshapes the Geometry of Distributed Semantic Representation.
Ontology highlight
ABSTRACT:
SUBMITTER: Nastase SA
PROVIDER: S-EPMC6248820 | biostudies-other | 2017 Aug
REPOSITORIES: biostudies-other
Publications
Attention Selectively Reshapes the Geometry of Distributed Semantic Representation.
Nastase Samuel A SA Connolly Andrew C AC Oosterhof Nikolaas N NN Halchenko Yaroslav O YO Guntupalli J Swaroop JS Visconti di Oleggio Castello Matteo M Gors Jason J Gobbini M Ida MI Haxby James V JV
Cerebral cortex (New York, N.Y. : 1991) 20170801 8
Humans prioritize different semantic qualities of a complex stimulus depending on their behavioral goals. These semantic features are encoded in distributed neural populations, yet it is unclear how attention might operate across these distributed representations. To address this, we presented participants with naturalistic video clips of animals behaving in their natural environments while the participants attended to either behavior or taxonomy. We used models of representational geometry to i ...[more]
Similar Datasets
Project description:Little is known about how attention changes the cortical representation of sensory information in humans. On the basis of neurophysiological evidence, we hypothesized that attention causes tuning changes to expand the representation of attended stimuli at the cost of unattended stimuli. To investigate this issue, we used functional magnetic resonance imaging to measure how semantic representation changed during visual search for different object categories in natural movies. We found that many voxels across occipito-temporal and fronto-parietal cortex shifted their tuning toward the attended category. These tuning shifts expanded the representation of the attended category and of semantically related, but unattended, categories, and compressed the representation of categories that were semantically dissimilar to the target. Attentional warping of semantic representation occurred even when the attended category was not present in the movie; thus, the effect was not a target-detection artifact. These results suggest that attention dynamically alters visual representation to optimize processing of behaviorally relevant objects during natural vision.
| S-EPMC3929490 | biostudies-literature
Project description:This article describes an unsupervised machine learning method for computing distributed vector representation of molecular fragments. These vectors encode fragment features in a continuous high-dimensional space and enable similarity computation between individual fragments, even for small fragments with only two heavy atoms. The method is based on a word embedding algorithm borrowed from natural language processing field, and approximately 6 million unlabeled PubChem chemicals were used for training. The resulting dense fragment vectors are in contrast to the traditional sparse "one-hot" fragment representation and capture rich relational structure in the fragment space. The vectors of small linear fragments were averaged to yield distributed vectors of bigger fragments and molecules, which were used for different tasks, e.g., clustering, ligand recall, and quantitative structure-activity relationship modeling. The distributed vectors were found to be better at clustering ring systems and recall of kinase ligands as compared to standard binary fingerprints. This work demonstrates unsupervised learning of fragment chemistry from large sets of unlabeled chemical structures and subsequent application to supervised training on relatively small data sets of labeled chemicals.
| S-EPMC6044751 | biostudies-literature
Project description:Healthy individuals display systematic inaccuracies when allocating attention to perceptual space. Under many conditions, optimized spatial attention processing of the right hemisphere's frontoparietal attention network directs more attention to the left side of perceptual space than the right. This is the pseudoneglect effect. We present evidence reshaping our fundamental understanding of this neural mechanism. We describe a previously unrecognized, but reliable, attention bias to the right side of perceptual space that is associated with semantic object processing. Using an object bisection task, we revealed a significant rightward bias distinct from the leftward bias elicited by the traditional line bisection task. In Experiment 2, object-like shapes that were not easily recognizable exhibited an attention bias between that of horizontal lines and objects. Our results support our proposal that the rightward attention bias is a product of semantic processing and its lateralization in the left hemisphere. In Experiment 3, our novel object-based adaptation of the landmark task further supported this proposition and revealed temporal dynamics of the effect. This research provides novel and crucial insight into the systems supporting intricate and complex attention allocation and provides impetus for a shift toward studying attention in ways that increasingly reflect our complex environments. (PsycInfo Database Record (c) 2021 APA, all rights reserved).
| S-EPMC7818672 | biostudies-literature
Project description:What are the defining features of lay people's semantic representation of risk? We contribute to mapping the semantics of risk based on word associations to provide insight into both universal and individual differences in the representation of risk. Specifically, we introduce a mini-snowball word association paradigm and use the tools of network and sentiment analysis to characterize the semantics of risk. We find that association-based representations not only corroborate but also extend those extracted from past survey- and text-based approaches. Crucially, we find that the semantics of risk show universal properties and individual and group differences. Most notably, while semantic clusters generalize across languages, their frequency varies systematically across demographic groups, with older and female respondents showing more negative connotations and mentioning more often certain types of activities (e.g., recreational activities) relative to younger adults and males, respectively. Our work has general implications for the measurement of risk-related constructs by suggesting that "risk" can mean different things to different individuals.
| S-EPMC9269897 | biostudies-literature
Project description:BackgroundEmbedding techniques for converting high-dimensional sparse data into low-dimensional distributed representations have been gaining popularity in various fields of research. In deep learning models, embedding is commonly used and proven to be more effective than naive binary representation. However, yet no attempt has been made to embed highly sparse mutation profiles into densely distributed representations. Since binary representation does not capture biological context, its use is limited in many applications such as discovering novel driver mutations. Additionally, training distributed representations of mutations is challenging due to a relatively small amount of available biological data compared with the large amount of text corpus data in text mining fields.MethodsWe introduce Mut2Vec, a novel computational pipeline that can be used to create a distributed representation of cancerous mutations. Mut2Vec is trained on cancer profiles using Skip-Gram since cancer can be characterized by a series of co-occurring mutations. We also augmented our pipeline with existing information in the biomedical literature and protein-protein interaction networks to compensate for the data insufficiency.ResultsTo evaluate our models, we conducted two experiments that involved the following tasks: a) visualizing driver and passenger mutations, b) identifying novel driver mutations using a clustering method. Our visualization showed a clear distinction between passenger mutations and driver mutations. We also found driver mutation candidates and proved that these were true driver mutations based on our literature survey. The pre-trained mutation vectors and the candidate driver mutations are publicly available at http://infos.korea.ac.kr/mut2vec .ConclusionsWe introduce Mut2Vec that can be utilized to generate distributed representations of mutations and experimentally validate the efficacy of the generated mutation representations. Mut2Vec can be used in various deep learning applications such as cancer classification and drug sensitivity prediction.
| S-EPMC5918431 | biostudies-literature
Project description:The ability for the brain to discriminate among visual stimuli is constrained by their retinal representations. Previous studies of visual discriminability have been limited to either low-dimensional artificial stimuli or pure theoretical considerations without a realistic encoding model. Here we propose a novel framework for understanding stimulus discriminability achieved by retinal representations of naturalistic stimuli with the method of information geometry. To model the joint probability distribution of neural responses conditioned on the stimulus, we created a stochastic encoding model of a population of salamander retinal ganglion cells based on a three-layer convolutional neural network model. This model not only accurately captured the mean response to natural scenes but also a variety of second-order statistics. With the model and the proposed theory, we computed the Fisher information metric over stimuli to study the most discriminable stimulus directions. We found that the most discriminable stimulus varied substantially across stimuli, allowing an examination of the relationship between the most discriminable stimulus and the current stimulus. By examining responses generated by the most discriminable stimuli we further found that the most discriminative response mode is often aligned with the most stochastic mode. This finding carries the important implication that under natural scenes, retinal noise correlations are information-limiting rather than increasing information transmission as has been previously speculated. We additionally observed that sensitivity saturates less in the population than for single cells and that as a function of firing rate, Fisher information varies less than sensitivity. We conclude that under natural scenes, population coding benefits from complementary coding and helps to equalize the information carried by different firing rates, which may facilitate decoding of the stimulus under principles of information maximization.
| S-EPMC10245665 | biostudies-literature
Project description:Accelerated by the Human Genome Project, genetic testing has become an increasingly integral component in diagnosis, treatment, management, and prevention of numerous diseases and conditions. More than 480 laboratories perform genetic tests for more than 4,600 rare and common medical conditions. These tests can effectively help health professionals to determine or predict the genetic conditions of their patients. However, physicians have not actively incorporated such innovative genetic technology into their clinical practices according to two national wide surveys commissioned by UnitedHealth Group. To fill the gap of insufficient use of a large number of genetic tests, we generated a single Resource Description Framework (RDF) resource, called GT2RDF (Genetic Testing data to RDF) by integrating information about disease, gene, phenotype, genetic test, and drug from multiple sources including Genetic Testing Registry (GTR), Online Mendelian Inheritance in Man (OMIM), MedGen, Human Phenotype Ontology (HPO), ClinVar, National Drug File Reference Terminology (NDF-RT). Meanwhile, we manually annotated and extracted information from 200 randomly selected GeneReviews chapters, and integrated into the GT2RDF. We performed two case studies to demonstrate the usability of the GT2RDF. GT2RDF will serve as a data foundation to support the design of a genetic testing recommendation system, called iGenetics, which will ultimately facilitate the pace of precision medicine by means of actively and effectively incorporating innovative genetic technology in clinical settings. Abbreviations: GT2RDF: Genetic Testing data to RDF; SWT: Semantic web technology; OWL: Ontology Web Language; RDF: Resource Description Framework; SPARQL: SPARQL Protocol and RDF Query Language; GTR: Genetic Testing Registry; OMIM: Online Mendelian Inheritance in Man; HPO: Human Phenotype Ontology; NDF-RT: National Drug File Reference Terminology; UMLS: Unified Medical Language System.
| S-EPMC5333271 | biostudies-literature
Project description:Object conceptual processing has been localized to distributed cortical regions that represent specific attributes. A challenging question is how object semantic space is formed. We tested a novel framework of representing semantic space in the pattern of white matter (WM) connections by extending the representational similarity analysis (RSA) to structural lesion pattern and behavioral data in 80 brain-damaged patients. For each WM connection, a neural representational dissimilarity matrix (RDM) was computed by first building machine-learning models with the voxel-wise WM lesion patterns as features to predict naming performance of a particular item and then computing the correlation between the predicted naming score and the actual naming score of another item in the testing patients. This correlation was used to build the neural RDM based on the assumption that if the connection pattern contains certain aspects of information shared by the naming processes of these two items, models trained with one item should also predict naming accuracy of the other. Correlating the neural RDM with various cognitive RDMs revealed that neural patterns in several WM connections that connect left occipital/middle temporal regions and anterior temporal regions associated with the object semantic space. Such associations were not attributable to modality-specific attributes (shape, manipulation, color, and motion), to peripheral picture-naming processes (picture visual similarity, phonological similarity), to broad semantic categories, or to the properties of the cortical regions that they connected, which tended to represent multiple modality-specific attributes. That is, the semantic space could be represented through WM connection patterns across cortical regions representing modality-specific attributes.
| S-EPMC5906027 | biostudies-literature
Project description:BACKGROUND: Digitised monogenean images are usually stored in file system directories in an unstructured manner. In this paper we propose a semantic representation of these images in the form of a Monogenean Haptoral Bar Image (MHBI) ontology, which are annotated with taxonomic classification, diagnostic hard part and image properties. The data we used are basically of the monogenean species found in fish, thus we built a simple Fish ontology to demonstrate how the host (fish) ontology can be linked to the MHBI ontology. This will enable linking of information from the monogenean ontology to the host species found in the fish ontology without changing the underlying schema for either of the ontologies. RESULTS: In this paper, we utilized the Taxonomic Data Working Group Life Sciences Identifier (TDWG LSID) vocabulary to represent our data and defined a new vocabulary which is specific for annotating monogenean haptoral bar images to develop the MHBI ontology and a merged MHBI-Fish ontologies. These ontologies are successfully evaluated using five criteria which are clarity, coherence, extendibility, ontology commitment and encoding bias. CONCLUSIONS: In this paper, we show that unstructured data can be represented in a structured form using semantics. In the process, we have come up with a new vocabulary for annotating the monogenean images with textual information. The proposed monogenean image ontology will form the basis of a monogenean knowledge base to assist researchers in retrieving information for their analysis.
| S-EPMC3639807 | biostudies-other
Project description:BackgroundReporting and sharing experimental metadata- such as the experimental design, characteristics of the samples, and procedures applied, along with the analysis results, in a standardised manner ensures that datasets are comprehensible and, in principle, reproducible, comparable and reusable. Furthermore, sharing datasets in formats designed for consumption by humans and machines will also maximize their use. The Investigation/Study/Assay (ISA) open source metadata tracking framework facilitates standards-compliant collection, curation, visualization, storage and sharing of datasets, leveraging on other platforms to enable analysis and publication. The ISA software suite includes several components used in increasingly diverse set of life science and biomedical domains; it is underpinned by a general-purpose format, ISA-Tab, and conversions exist into formats required by public repositories. While ISA-Tab works well mainly as a human readable format, we have also implemented a linked data approach to semantically define the ISA-Tab syntax.ResultsWe present a semantic web representation of the ISA-Tab syntax that complements ISA-Tab's syntactic interoperability with semantic interoperability. We introduce the linkedISA conversion tool from ISA-Tab to the Resource Description Framework (RDF), supporting mappings from the ISA syntax to multiple community-defined, open ontologies and capitalising on user-provided ontology annotations in the experimental metadata. We describe insights of the implementation and how annotations can be expanded driven by the metadata. We applied the conversion tool as part of Bio-GraphIIn, a web-based application supporting integration of the semantically-rich experimental descriptions. Designed in a user-friendly manner, the Bio-GraphIIn interface hides most of the complexities to the users, exposing a familiar tabular view of the experimental description to allow seamless interaction with the RDF representation, and visualising descriptors to drive the query over the semantic representation of the experimental design. In addition, we defined queries over the linkedISA RDF representation and demonstrated its use over the linkedISA conversion of datasets from Nature' Scientific Data online publication.ConclusionsOur linked data approach has allowed us to: 1) make the ISA-Tab semantics explicit and machine-processable, 2) exploit the existing ontology-based annotations in the ISA-Tab experimental descriptions, 3) augment the ISA-Tab syntax with new descriptive elements, 4) visualise and query elements related to the experimental design. Reasoning over ISA-Tab metadata and associated data will facilitate data integration and knowledge discovery.
| S-EPMC4255742 | biostudies-literature