Speaker: Tieliu Shi , PhD
              Professor of Bioinformatics & Computational Biology
              School of Life Sciences
              East China Normal University
              Email: tieliushi01@gmail.com
              Web: http://www.biomed.ecnu.edu.cn/biomeden/98/9d/c9349a104605/page.htm 

Time: 10:00-11:30 am , Jan. 11（Thursday） 

Venue: Room 300, SIBS Main Building, Yueyang Road 320 

Host: Prof. Sijia Wang
         CAS-MPG Partner Institute for Computational Biology 

Title: Standardized Disease Annotation Knowledge Bases for Precision Medicine 

Abstract:
Precision medicine for disease prevention and treatment strategiest is currently the most popular practice in clinical application. However, these promising applications greatly rely on data mining, sharing and exchange. Currently, one of the major challenges for the application of big data is the lack of uniformly structured data in clinical practice, because different Healthy Information Systems and different Electronic Medical Records are applied by different hospitals, and different vocabularies are used to describe clinical observations and treatment strategies by different healthcare providers. The high disparity in clinical data among hospitals and healthcare providers is a significant obstacle for the data sharing and downstream analysis. To facilitate precision medicine in clinical applications, it is necessary to build a platform to standardize clinical information and use it for the integration of various clinical resources for individual patient’ precise diagnosis, prognosis, and therapeutic strategies. 

To this end, we built systems to standardize and classify pediatric and rare diseases - PedAM (Pediatrics Annotation & Medicine) and eRAM (Encyclopedia of Rare Disease Annotation for Precision Medicine). Both platforms integrate biomedical resources and clinical data from Electronic Medical Records (EMRs) and to support the development of computational tools which will enable robust data analysis and integration. Currently, near 10 million abstracts from PubMed and 1 million full text articles from PubMed Central have been text-mined, the extracted sentences describing the disease-manifestation (D-M) and disease-phenotype (D-P) are all stored in our database and can be traced back to the original published papers through PubMed ID. 

Currently, eRAM contains 15,942 diseases, with 27,329 unified human phenotype terms, 12,207 phenotypes from matched corresponding mouse phenotype ontology, 75,335 manifestations, when PedAM contains standardized 8528 pediatric disease terms (4542 unique disease concepts and 3986 synonyms). 

All are welcome, and anyone who would like to have a talk with speaker, please feel free to contact Jing MU via mujing@picb.ac.cn.