Although many data mining companies are competing for customers in the life sciences these days, newcomers are still confident they will find their own niche — but not always with tools alone. In a recent example of the expanding field, Phase-it Intelligent Solutions, a data mining company founded in Heidelberg, Germany, in late 2000, entered a service partnership with Mice & More. Over the coming months, Phase-it will work with the Hamburg-based transgenic mouse startup to integrate and analyze mass spectrometry, histopathology, and genotypic and phenotypic data from a breast cancer mouse model in order to develop new diagnostic tools.
Though it’s Phase-it’s first go at proteomics data, “they are quite experienced in this kind of analysis of tumor material, especially from a biological point of view,” commented Christine Schulze-Garg, co-founder and managing partner of Mice & More. This is precisely what Phase-it sees as its competitive edge: Founded by two former Lion Bioscience employees and two researchers from the German Cancer Research Institute (DKFZ), the company said it understands both the biological and the IT dimensions of a customer’s project. “We don’t need an interpreter to talk to each other, we understand the ‘native biologist’s’ language,” said Peter Wiesner, Phase-it’s CEO and a former vice president as well as the seventh employee at Lion. This, he believes, is crucial: “We have seen several corporations fail because the scientists could not really communicate with the IT people efficiently.”
According to Wiesner, Phase-it does not mainly distinguish itself by its algorithms but by the way it links them in analysis processes to create classifiers that can answer a customer’s question. It uses a large number of supervised and unsupervised machine-learning methods such as support vector machines, self-organizing maps, artificial neural networks, and Bayesian approaches. Some of these algorithms are publicly available, Wiesner said, others were custom-designed or modified by Phase-it. “We are not attempting to become a tool company like most of the other data mining companies,” he said.
Phase-it’s hope is to turn service partnerships like the one with Mice & More into more integrated collaborations that include downstream rights to inventions or product revenues. “We are a young company, new on the market, so of course we have to first demonstrate what we claim that we are capable of,” explained Wiesner. In fact, a successful feasibility study with DKFZ, the University of Ulm, and Boehringer Ingelheim encouraged the founding of the company. Wiesner said that several other pilot projects are on their way, including a toxicogenomics project with an undisclosed large pharmaceutical company.
Phase-it, which currently has 12 employees, is funded privately and through a research grant from the German government, but is looking into a larger project involving tumor markers that would require raising more money, most probably venture capital. Lion, it seems, is in no danger of being eaten by its roaring cub: “Our business is so completely different,” said Wiesner, since it focuses on solutions for specific projects rather than providing IT infrastructure. But his former employer remains an inspiration: “The success story of Lion really motivated us to try it ourselves.”