About this Research Topic
Life science knowledge is complex, large-scale and heterogeneous; causing information overload for many researchers and an interesting set of computational challenges for informatics systems designers, developers and curators. The published literature is growing exponentially providing a textual corpus consisting of over twenty million documents. Publications’ supplemental data have few enforced formatting or semantic standards that support effective data sharing. Finally, there are increasing numbers of biomedical informatics resources becoming available (databases, ontologies, controlled vocabularies, etc.), each serving a different purpose and increasing the level of heterogeneity within the data landscape whilst also providing valuable diversity of function for users.
The overload problem is particularly important in neuroscience (and neuroinformatics) due to the subject’s interdisciplinary nature involving multiple subfields: molecular and cellular biology, genetics, neurophysiology, clinical imaging, neuroanatomy, computer modeling, psychology, artificial intelligence, etc.. We here ask how computational methods can be used across these available resources (the literature, online databases, laboratory data) to alleviate the impact of overload, leverage the available diversity of resources and ultimately facilitate discovery?
This research topic aims to tackle this challenge in neuroscience by exploring novel approaches to computational capture and utilization of emergent semantics implied by biomedical texts and data. The approaches should be explicitly motivated by facilitation of discoveries in neuroscience by means of automated or semi-automated formulation and verification of scientific hypotheses and models. As a non-exclusive list, we expect submissions to involve (A) identification of important concepts in neuroscience literature and data; (B) automated interlinking of the concepts through lexical analysis, or the use of formal semantics; (C) identification of complex domain-specific relationships beyond mere similarities; (D) utilizing relationships between concepts to formulate machine-readable and actionable research questions, (E) automated verification of questions and intuitive presentation of the results to neuroscientists, etc.
We encourage contributions especially (but not only) from the following fields applied to neuroscience information :
• text mining,
• question answering,
• distributional semantics,
• formal semantics,
• ontology engineering,
• biomedical knowledge modeling,
• information integration,
• graph analysis,
• analytical data processing,
• data visualisation,
• intelligent user interfaces.
Submissions may combine approaches from multiple disciplines and should demonstrate practical applicability to the formulation and verification of hypothesis in neuroscience.
The overload problem is particularly important in neuroscience (and neuroinformatics) due to the subject’s interdisciplinary nature involving multiple subfields: molecular and cellular biology, genetics, neurophysiology, clinical imaging, neuroanatomy, computer modeling, psychology, artificial intelligence, etc.. We here ask how computational methods can be used across these available resources (the literature, online databases, laboratory data) to alleviate the impact of overload, leverage the available diversity of resources and ultimately facilitate discovery?
This research topic aims to tackle this challenge in neuroscience by exploring novel approaches to computational capture and utilization of emergent semantics implied by biomedical texts and data. The approaches should be explicitly motivated by facilitation of discoveries in neuroscience by means of automated or semi-automated formulation and verification of scientific hypotheses and models. As a non-exclusive list, we expect submissions to involve (A) identification of important concepts in neuroscience literature and data; (B) automated interlinking of the concepts through lexical analysis, or the use of formal semantics; (C) identification of complex domain-specific relationships beyond mere similarities; (D) utilizing relationships between concepts to formulate machine-readable and actionable research questions, (E) automated verification of questions and intuitive presentation of the results to neuroscientists, etc.
We encourage contributions especially (but not only) from the following fields applied to neuroscience information :
• text mining,
• question answering,
• distributional semantics,
• formal semantics,
• ontology engineering,
• biomedical knowledge modeling,
• information integration,
• graph analysis,
• analytical data processing,
• data visualisation,
• intelligent user interfaces.
Submissions may combine approaches from multiple disciplines and should demonstrate practical applicability to the formulation and verification of hypothesis in neuroscience.
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.
