BACKGROUND: BioVisioN is an independent, venture capital- backed, biotechnology R&D company. The mission of BioVisioN is the development and commercialization of technologies for applications in drug discovery and diagnostics. BioVisioN's powerful technologies include high throughput identification of peptides and proteins, resulting in the generation of large databases. BioVisioN uses Differential Peptide Display for developments in neurodegenerative diseases, cancer and bone diseases. Access to our technologies is preferably provided through collaborative alliances. The company was founded in 1997 as a spin-off of an academic research institute, the Lower Saxony Institute for Peptide Research (IPF), and is located in close vicinity to Hanover Medical School in Germany.

PLATFORMS: Megaputer PolyAnalyst Power 4.1 runs on any typical Windows platform. The more powerful version, PolyAnalyst Pro, requires Windows NT as an operating system. We use it with NT 4.0 (Service pack 5) and 128MB RAM.

PROBLEM SOLVED: The progress of the biological science, due to the human genome project and the progress of measurement technology applied in biochemistry, produces very large quantities of data. Peptides (i.e., the molecule insulin) contribute to many important biochemical processes within the human body. There are thousands of different molecules, and BioVisioN tries to find relationships to disease(s) by analyzing their complex patterns in human body fluids such as blood and urine. For the finding of biomarkers, and for the development of diagnostics, it is important to find the most striking heuristics for the prediction or discrimination of disease. PolyAnalyst is one of BioVisioN's software tools to address this problem.

PRODUCT FUNCTIONALITY: Due to the integrated analysis environment, PolyAnalyst can be used intuitively, inviting experimentation with different exploration engines for finding models, discrimination or classification. There are different versions of PolyAnalyst. The Lite version offers machine learning algorithms: find dependencies, linear regression, classify, discriminate, summary statistics and find clusters. The more advanced versions are extended by the find laws, poly net predictor (a GMDH derivative), memory-based reasoning and market basket analysis. Any of the exploration engines, including the find laws engine, can be obtained as individual COM components. These COM objects may be integrated into other software projects outside of the PolyAnalyst environment, which BioVisioN plans to do.

STRENGTHS: The product is easy to use, and most users require only a few analysis sessions using all the offered tools to become proficient. The program's exploration engines work fast, and intermediate results are offered continuously during the exploration. This allows supervision of the model-forming process. The exploration engines appear very sophisticated, and they really do their job. It is able to efficiently handle large data sets with many cases.

WEAKNESSES: The basic statistical functions of PolyAnalyst should become even more powerful. PolyAnalyst is, of course, focused on sophisticated exploration engines, but this improvement would be of great value in order to reduce the time spent using standard statistical packages. Furthermore, the extension of the cluster engine would be on my wish list for PolyAnalyst, which means the inclusion of hierarchical clustering methods.

SELECTION CRITERIA: One of the main selection criteria was the possibility of testing the software before purchasing it, which in the case of PolyAnalyst can be easily achieved by downloading an evaluation copy from the Web. The Lite version of the system is inexpensive and allows getting more familiar with the concepts of PolyAnalyst. This allowed us to gradually learn the product and upgrade the system to the more expensive and more sophisticated members of the product line step by step.

DELIVERABLES: PolyAnalyst provides different reporting options, such as graphs or tables for the description of clusters. More importantly, most PolyAnalyst exploration engines produce equations describing the underlining heuristics. The output is bundled in a compact report sheet. The combination of raw data, rules, graphs and models is very well organized with intuitive Windows GUI elements.

VENDOR SUPPORT: It is easy to contact Megaputer Intelligence Inc. by e-mail. The support personnel always appeared friendly and enthusiastic about serving our needs.

DOCUMENTATION: The documentation we used is present on the Megaputer Web site. The tutorials, provided by PolyAnalyst, are exceptionally useful for gaining insight in the step-by-step analysis process.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access