Waters, Umetrics simplify drug discovery stats
Waters Corp. announced recently that it had entered into a co-marketing agreement with Umetrics AB of Umea, Sweden, under which Waters will begin offering scientists a direct link to Umetrics’ SIMCA-P multivariate statistical analysis software as an option compatible with Waters’ MarkerLynx 4.1 Software.
Register for free to listen to this article
Listen with Speechify
0:00
5:00
SEATTLE—Waters Corp. announced recently that it had entered into a co-marketing agreement with Umetrics AB of Umea, Sweden, under which Waters will begin offering scientists a direct link to Umetrics' SIMCA-P multivariate statistical analysis software as an option compatible with Waters' MarkerLynx 4.1 Software.
This is only the first step in a series of projects to further integrate the companies' products, according to Dr. John Shockcor, business development manager for metabolic profiling at Waters. "It's not 'the' integration but rather 'an' integration of our products, which allows data to move seamlessly between the two packages and give end users the ability to more easily use powerful statistical tools to mine data."
SIMCA-P provides users of data analyzed by MarkerLynx software with a suite of statistical tools that include partial least squares discriminate analysis and orthogonal partial least squares. These advanced statistical tools, along with a complete graphical display suite, allow researchers to visualize their data and extract more meaningful information from their experimental results, Waters reports.
But the end result is to keep the end users from having to think about all that.
"We want to keep users from having to get too deeply into the statistical portion of things themselves," Shockcor says. "We want to take the mumbo-jumbo out of analysis. One of the problems with some of these advanced lab techniques and tools is that people get intimidated by the word 'statistics.'"
And that intimidation all goes back to Mark Twain's quote that there are three types of lies: lies, damned lies and statistics.
"Users know how easy it can be to do something mistakenly that, for example, gives you the result you want as opposed to what the result really is," Shockcor says. "Or that makes you think you aren't getting what you want when indeed you are. We're all about using this collaboration to minimize the difficulties in analysis and reduce the chance for error."
He also notes that anytime drug discovery researchers analyze complex mixtures or compounds, there can be ambiguity as to whether there is a true difference between two or more groups that are being compared. With this first step in integrating the Waters and Umetrics products—and future integration work as well—users will reportedly be able to have strong confidence that when the analysis says there is a difference, that it is indeed a significant one.
"Modern analytical instrumentation generates large amounts of information-rich data," notes Svante Wold, VP of research and development for Umetrics. "Combined with the powerful chemometrics data analytical methodology implemented in Umetrics' SIMCA-P software, the scientist is provided with a broad approach for investigating the properties of complex samples from processes in the pharmaceutical, biotech, and neighboring areas. Integrated and graphically oriented solutions like the combination of Waters and Umetrics software make it faster to pursue, visualize and interpret the results from these investigations."
And what of the future integration work between Waters and Umetrics? Shockcor is keeping mostly mum on what surprises are yet in store ("I have competitors out there and don't want to tip my hand too much"), but he does say that as the products become more integrated, the process of analyzing statistics will become more and more automatic and seamless. Although he hesitates to use the word "wizard," he does suggest that the process of crunching numbers will become much more like working through a wizard-type interface.
"We want to make it possible for even the techs who aren't as scientifically or statistically trained to be able to do a fair amount of work, with minimal supervision, to generate useful, accurate and reliable data using SIMCA-P and MarkerLynx," he explains. "We will never be offering a 'black box,' but we do want it to be something that won't be intimidating for the scientists or their assistants."