SYNTHETIC VIEWS TO TRADITIONAL HERBAL MEDICINE DERIVED FROM TEXT MINING OF BIOMEDICAL BIBLIOGRAPHIES

Abstract

Text mining approach was applied to a subset of MEDLINE database in a conjunction with several chemoinformatics and bioinformatics databases. The connections between biomedical herbs, active compounds, traditional cultures and most important human diseases were collected. The integrated statistical properties of the obtained connections were estimated. The consistency of the obtained distributions with a conventional approaches of molecular biology and medicine was approved.

The generic statistical properties of the obtained connections allow to clarify and specify the generic relationships between herbal cultures and between diseases of humans. The averaged content of chemical compounds was in agreement with abundances of specific enzymes in plant genomes. The role of traditional Chinese medicine as most stable and conservative herbal culture was approved.

The clear separation of mental disorders from oncology and cardiovascular diseases was observed. A uniform activity of plants associated with a treatment of oncology diseases appears to be in contrast to a diverse activity of plants used to correct brain disorders. The obtained results could be meaningful to close the gap between traditional and modern approaches of medicine.