The advent of inexpensive, powerful, personal computers has been a boon for chromatography data systems. Today, all instruments ship with software packages that control instrumentation, acquire data, prepare reports, manipulate spectra and chromatograms, store methods, perform calibrations, and interface with lab- or organization-wide computer systems. This was not always so. Before PCs were commonly available, chromatographs shipped with integrators and, before that, strip chart recorders.
Yet the fundamental task of data systems remains the same. “Their basic function,” says Hugh Goldsmith, president of GC systems manufacturer SRI (Torrance, CA), “is to stripchart the data, integrate the peaks, and calibrate them so your answer appears in meaningful units.” Goldsmith recalls that in the “old days” users integrated peaks by physically cutting out traces and weighing them on a balance. Shoulders and overlapping peaks were dealt with by handdrawing, or by extrapolating the peak of interest to baseline. “Now you can dice and slice the chromatogram electronically because you can always call up the original digital representation. You don’t even need paper.”
Division of labor
According to Jim Edwards, business development manager at Thermo Fisher Scientific (Austin, TX), chromatography data systems are becoming more versatile and less instrumentcentric. “We increasingly find that as labs grow and acquire GCs from different vendors, more software systems are capable of controlling third-party instruments. Several software-only firms, for example, establish relationships with companies such as ours to develop drivers and controls for our instruments that operate natively within their software. Years ago that would have been difficult to do because of the costs of computer systems.”
Improvements in chromatography data systems have for the most part tracked advances in hardware—a very desirable development given the tremendous strides in chromatographic methods and instruments. “If all vendors did was focus on improving hardware to make it more productive and less noisy, we’d simply have a bottleneck on the software side,” notes Edwards.
Data system designers have quite efficiently exploited the doubling of computing power every few years and falling prices of computer systems. Software now guides users through routine operations, method development, and basic report generation, leaving scientists and technicians—the real “CPUs” of a chromatography system—to focus on designing experiments, manipulating and reducing data, and reviewing results. “It’s been a productive division of labor,” Edwards says.
It’s all in the interface
Chromatography data systems have had to adapt to the slowly evolving demographic of laboratory personnel, from specialist to generalist. “We have observed a shift from users who were chromatographers and analytical chemists to a new group composed of process manager, quality manager, statistician, biologist, and so on,” observes Alessandro Baldi, business manager for MS and chromatography software at PerkinElmer (Waltham, MA). The change does not necessarily represent diminishing expertise, although that is certainly one component of it. But it does reflect the prevailing sentiment among both analysts and equipment suppliers that instruments are tools for rather than objects of laboratory work. Chromatographs need to be intuitive and easy to use, and a key component of usability is the interface. Baldi cites popular gadgets and software whose interfaces have become simple to the point where individuals can operate them with no training.
Can this be achieved with analytical instruments, and if so, is it desirable? Baldi answers “yes” to both questions. PerkinElmer has put a lot of effort into its graphical user interface, to provide simplicity and ease of use that Baldi says have become his company’s mantras. Software developers, he argues, should exploit the familiarity of icons and visual components from consumer-grade software, for example, to initiate or end a run, and structure the navigation in their data systems similarly to common e-mail clients. “Data systems these days resemble iTunes or Outlook more than they do the chromatography system of 20 years ago.”
While achieving this level of userfriendliness, data system designers must somehow cope with the everincreasing intricacy of analysis. Many validated methods, particularly those supporting regulated industries or law enforcement, involve the use of multiple columns and detectors, which generate outputs in multiple dimensions. PerkinElmer deals with this complexity by creating layers similar to those in popular software. “You don’t see everything at once, but instead have intuitive navigation from one environment to another, just as you would within programs that bundle e-mail, contacts, and calendar functions. After you navigate to the different layers, it becomes horizontal and easier to visualize. This is the kind of architecture we have been trying to mimic.”
Another strategy for improving the user experience is to improve the data product continually. PerkinElmer’s Chromera software package, released at Pittcon 2009, undergoes new releases quarterly. Like most instrument vendor packages, which work only with the respective vendor’s instruments, Chromera works only with PerkinElmer instruments, but the eventual goal is interoperability with third-party instruments.
Angelo DePalma holds a Ph.D. in organic chemistry and has worked in the pharmaceutical industry. You can reach him at angelo@ adepalma.com.