HSLS Update

CellSpace Knowledge Miner Available Through HSLS

In recent years, burgeoning research in the field of life sciences has produced massive information overload. With the development of bibliographic databases, information retrieval is relatively easy, but capturing knowledge from the search results has become increasingly difficult. For example, if you want to know the major proteins or gene products involved in the biological process of apoptosis (cell death), a PubMed search of “apoptosis” will return a list of roughly 69,000 articles for the last 30 years. Deriving knowledge from this overwhelming amount of data is difficult and time consuming. In an effort to enhance the knowledge capture process from PubMed PubMed, HSLS has licensed CellSpace Knowledge Miner from Cellomics Inc. By using this tool in the previous example, a searcher would be able to find a list of proteins that co-occur with the term “apoptosis” in the research literature. In general, co-occurring terms have a biologically meaningful relationship.

CellSpace is a bioinformatics tool— a knowledge mining system that automatically detects, analyzes, and reports the logical relationships between four types of terms found in the research literature:

1. molecule: proteins, genes, drugs
2. function: biological processes and disease states
3. cell type
4. organism

CellSpace computers analyze the National Library of Medicine’s MEDLINE database, performing proprietary statistical correlation analyses regarding the organisms, cell types, biological processes, and molecules reported in 655 selected life science research journals. The molecular relationships extracted from the literature are then stored in the CellSpace database, which can be queried via the CellSpace user interface. The information is updated every two weeks.

CellSpace allows users to start with one or more molecules, functions, cells and systems, or organisms, and search for connections in any of the four categories. Using CellSpace, searchers can:

• Start with a single protein (or other molecule) and find its functions, the diseases in which it is implicated, and related molecules.
• Start with results from a high-throughput experiment (such as a cluster of co-regulated genes from microarray analysis), and easily find the functions that they share.
• Start with the results of proteomics experiments, and quickly screen the data to distinguish published interactions from novel ones.
• Start with a disease and find related molecules, or related functions.
• Start with two or more functions, and find the related molecules that they have in common.
• View the literature that supports the connections found in CellSpace.

CellSpace is available via the HSLS molecular biology and genetics resource <www.hsls.pitt.edu/guides/genetics> listed under “Resource Highlights”. Access is limited to 8 simultaneous PC users (Mac OS is not supported) from UPMC and the University of Pittsburgh Oakland campus. For more information or to schedule CellSpace training please contact Ansuman Chattopadhyay, PhD, information specialist in molecular biology and genetics, at 412-648-1297 or ansuman@pitt.edu.

--Ansuman Chattopadhyay