SciNova Informatics, a Pune based startup has launched Prometheus -- a modular, cutting edge yet user-friendly data mining product for Life Science Research. It enables researchers to use their domain expertise and not bother with the underlying complexity of data mining techniques.

Talking about the product, CSO Rajeev Gangal said, "Life Science data from the molecular to the clinical level is typically multivariate and inherently non-linear. Prometheus has algorithms like SVM's, Neural Networks and Decision Trees that can build accurate predictive models to hasten research." He further added Prometheus helps in trying out multiple machine learning algorithms and decide which is the best. A combination of these tools in series or parallel can also be used to get the most accurate result.

Further dwelling on the usage of the product Gangal said Prometheus could be used for mining of proteins, DNA and small organic molecules and other data. It is completely platform independent, since it has been developed in Java. "Prometheus is designed to give maximum flexibility to the user," said he. Prometheus uses a visual pipeline approach for data mining. The user can enter or exit the pipeline at any stage and store the process for future reference. The whole project can be reloaded whenever required explained Gangal.

Prometheus can be used in medical informatics, too, wherein in a typical hospital environment, it can predict a patient's length of stay (LOS), the tests required based on a combination of demographic, diagnostic and clinical variables. "When calculated for over a year, it can save loads of money for both, the hospital administration and patients," stated Gangal. Amongst other applications it can be used for eukaryotic promoter prediction, which is one of the most attractive problems in sequence analysis today. Promoter prediction is required to understand the intricacies of regulation and expression of many uncharacterized genes, and to determine functionality of many useful genes. The other applications include protein fold prediction, QSAR, target identification, structure elucidation of natural compounds.

Under its marketing plans, SciNova primarily aims to tap the overseas market. Said Gangal, "We are trying to tap the South- East Asian market. We aim to have a major market share from the overseas market and we are already in talks with few overseas companies."