[Obtaining new information on hemolytic uremic syndrome by text mining]

Area of Science:

• Biomedical Informatics
• Data Science in Medicine
• Computational Biology

Background:

• Hemolytic Uremic Syndrome (HUS) is a severe condition characterized by thrombotic microangiopathy, hemolytic anemia, thrombocytopenia, and acute renal failure, often leading to severe complications or death, particularly in children.
• Existing research on HUS is extensive, necessitating advanced methods to synthesize information and identify novel insights.

Purpose of the Study:

• To analyze explicit and implicit information within 16,192 scientific articles on HUS using text mining.
• To identify undescribed associations, extract underlying data, and forecast research trends related to HUS.
• To develop and apply computational workflows for data science in HUS research.

Main Methods:

• Text mining (TM) techniques were applied to abstracts of 16,192 HUS articles from the Europe PMC database.
• Specially developed workflows (WF) on the KNIME platform were utilized for data analysis.
• Unsupervised algorithms were employed for thematic clustering and trend forecasting.

Main Results:

• Undescribed associations between HUS-related events were detected.
• Underlying information and thematic clusters within HUS research were extracted.
• Forecasting of future research directions in HUS was performed.

Conclusions:

• Text mining of HUS literature provides valuable insights into research behaviors and trends.
• The developed data science approach and workflows are transferable to other biomedical topics and databases.
• This methodology can improve the research, prevention, and treatment of various human diseases.