From Plants to Human Health: Bridging the Gap with Knowledge Based Discovery

Diet related chronic diseases have reached epidemic proportions, prompting researchers to investigate the complex effects diet has on human health. Fruit and vegetable rich diets have been associated with health benefits but the molecular mechanisms describing these associations remain largely unknown. The scientific literature contains associations of bioactive components in fruits and vegetables and their interaction with human genes and gene products. The literature spans a multitude of life science domains and presents a massive volume of data for mining.

Identifying descriptive associations from scientific literature between dietary components and their molecular interactions with the human genome requires high throughput text-mining methods to process massive volumes of free text.

An NLP-based software package was used to extract semantic relationships to explain the molecular mechanisms between diet and human health. Text-mined results were integrated with relationships extracted from curated ontologies and association databases from NCBI, USDA, and EMBL-EBI to generate a high-density graph database.

This study produced a graph-based semantic network of dietary components and disease related genes to assist in the elucidation of health benefits conferred by plant-based foods. This work provides a foundational resource for further research in nutritional genomics to discover bioactive compounds for disease management through personalized nutrition.