ECO-CollecTF

ECO-CollecTF annotates evidence-based assertions in biomedical manuscripts to standardize descriptions of genes, gene products, and other biological entities derived from high-throughput experiments.

Key Features:

• Evidence-Based Annotation: Annotates statements that are directly supported by evidence presented in biomedical journal articles, linking assertions to the specific findings they support.
• Use of Evidence and Conclusion Ontology (ECO): Employs the Evidence and Conclusion Ontology (ECO) to structure and categorize evidence types for consistent annotation.
• High-Quality Corpus: Provides a curated collection of 84 documents containing annotated evidence-based statements for development and evaluation of text mining methods.

Scientific Applications:

• Text Mining Tool Development: Serves as a standardized, evidence-annotated corpus for developing and tuning text-mining algorithms that extract assertions about genes and gene products.
• Reliability Assessment: Supports assessment of the reliability of scientific assertions by linking statements directly to their supporting experimental evidence.

Methodology:

Expert curation and Evidence and Conclusion Ontology (ECO)-based annotation were applied to identify, extract, and harmonize evidence-based statements across the corpus.