Causality Knowledge Models
We develop biological network models which Capture, Represent and Analyze knowledge in biological sciences
Advancements in high-throughput technologies and diagnostic techniques have produced an abundance of data in biological sciences. However, putting data into context for domain specific applications- the extraction of knowledge - remains a challenge unsolved. As we have a limited number of approved drugs for most complex idiopathic diseases, the life sciences industry has long acknowledged the need to capture knowledge and represent it in computable form. Adoption of both human and machine -readable biological network models as the new semantic formalism to capture and present knowledge in life sciences helps to make sense out of the overtly complex ‘data-chaos’.
Our specialized set of tools are capable of transforming scientific findings in biological context into computable formats, giving rise to our 'causality biomodels'. These biomodels, along with capturing the causal and correlative aspects of biological entities and corresponding relationships, also represent the extracted knowledge with sufficient background information including the experimental setup, citation, supporting lines of evidence and other context information.
Specifications of extraction are:
Abstract, Materials and Methods (only for Annotation extraction), and Results.
Gene/protein entities including details on gene variants and protein modification and on the molecular, cellular and systemic effects of drugs and chemicals.
Causal (increase, decrease) and Correlation (positive, Negative and Association, Regulates).
Name, Version, Description, Copyright, Authors, Licenses, Contact Information.
Citation (e.g., PubMedID, DOI, title) and Support (sentence(s), table row, figure).
Species, Disease, CellLine, Cellular Location, Experimental system (e.g., IPS cells, mouse model, cell line), Experimental set-up (e.g., assay, sampling)
We develop biological network models which capture, represent and analyze knowledge in biological sciences. Our models have computable data (including, but not limited to transcriptomics, proteomics/phosphoproteomics, metabolomics data) extracted from various sources including biomedical literature, databases, books and patents, Electronic Health records etc., sorted and presented with formal syntax, which can be operated upon by algorithms to find valuable insights. The complex semantic networks of relationships in our knowledge models allow one to detect connections, extract patterns and answer complex research questions. Thus, they lay the ground for the identification of causalities of diseases through analysis of their causative mechanisms and hence are aptly named as "Causality knowledge models".