Molecular Connections’ meta-data extraction service is highly efficient as it employs a hybrid workflow combining automated data extraction followed by manual curation by the subject matter experts. The level of automation can be decided upon the nature of the information sources (both structured and unstructured sources) and in consultation with the client to improve efficiency. The workflow is customizable and can handle specific client requirements across sources and domain.
Molecular Connections expertise in data mining/ Meta-data creation spans across various vertices including collection, validation, enrichment, normalization. Our services include:
Meta – data creation and enhancement
Data normalization and linking
Semantic Fingerprinting & Enrichment
Automated Indexing Services
Document abstraction and Indexing
Inline Tagging / Indexing
Multimedia, Image and Indexing
The service also helps you deal with complexities in big-data normalization by attributing named entities to identifiers either by recognizing pre-defined standard identifiers or domain specific heuristic identifiers. The service also helps you deal with context specificity of the indexed ontology terms within the scope of how it is defined in the ontology.
The indexing and abstraction capabilities offered by this service are routinely used by the clients to index and create abstracts for a variety of documents such as patents, technical literature, scientific literature, journals , books, news articles. Molecular Connections’ semantic indexing capabilities identify key concepts and entities within the text sources and they are further indexed using standardized vocabularies.
Proprietary Text mining System – MC Miner™
The system is equipped with several proprietary algorithms and methods to deal with the complexities of real-world data. The system facilitates employing an ensemble of both rule based and a statistical method to fine tune the information extraction process.
Select Case Studies
Understanding Pharmacologic actions, protein interactions and toxicity studies using XTractorTM.
In this case study we extracted the Pharmacological actions, protein interactions and toxicity of various drugs and we also manually curated the side-effects from the drug labels
We undertook a case study to identify the multiple target preferences and polypharmacology of a select set of drugs using the XTractorTM platform to identify the repurposing potential.