|
Case Study: Duke University MedicalOur Story of Using Harvest™ Software Dr. Shengchu Wang and I are biostatisticians in Duke University Medical School. We are interested in bioinformatics research in medical genetics and genomics. We have been working on a gigantic amount of genomics and transcriptomics data and navigating in the sea of biomedical information. We have used Harvest™ in our research since October 2008. Comparing with many other bioinformatics software solutions, Harvest™ is unique in that it provides logical inference in the representation of knowledge for a large number of biochemical and medical gene pathways, collected from public databases, such as KEGG and BioGRID. By saying "logic inferences", we mean (1) Harvest™ has developed a smart approach to generate "new" knowledge based on the "old" knowledge, with user-defined "rules". Comparing with some popular text-mining softwares, such as Ingenuity®, Gene2network® and Pathway Studio®, which are more dictionary-like, Harvest™ can "learn" and "think", more like the human mind. When knowledge is abundant and "rules" are clearly defined and appropriate, Harvest™ can discover more; (2) Harvest™ can easily accumulate and manage knowledge that may be efficiently reused. This accumulated knowledge, or "templates", will be reused to collect more and more knowledge quickly and thoroughly for different bio-med and/or bioinformatics domains. For many other bioinformatics software solutions, often the computing time increases exponentially as the size of the searching domain and the number of types of correlations (rules) increase linearly. We did not note this when using Harvest™. The computing speed is fast, even with the logical inference and a large number of pathways. It was explained to us that this is achieved by using a real-time algorithm in their inference engine. This allows us to dynamically search and analyze a large number of pathways in real-time. Because of this, we are able to save a lot of time in our research. Finally, we really like Aedi Bio's customer support. They have customized the Harvest™ application to our specified needs in a very timely manner. We will recommend Harvest™ to a couple of other Duke Medical research labs. Sheng Feng, Ph.D. August 20, 2009 [Click here to view letter in PDF form]
|