Advancing bioprocessing depends as much—sometimes more—on data analysis as it does on analytical techniques or manufacturing processes. In some cases, improved analysis arises from the development of in-house techniques (see Solving the Sequence-Variant Problem in Therapeutic Protein Manufacture), but long-standing tools can also be used. As one example, Zayla Schaeffer, associate scientist AstraZeneca in Gaithersburg Maryland, and her colleagues used MATLAB to analyze glycosylation.
As the Schaeffer team noted: “The glycosylation profile of a therapeutic protein can impact its half-life, efficacy, and interactions with the human body.” Nonetheless, the range of glycan structures makes it extremely difficult to characterize all the glycan sites in a biotherapeutic protein.
Typically, scientists characterize these sites with liquid chromatography (LC) followed by mass spectrometry (MS). Although the resulting data can be analyzed manually, Schaeffer and her colleagues point outed that this “can be time-consuming and arduous.” Even where software tools exist to analyze LC-MS or LC-MS/MS data from glycan analysis, these scientists indicated that high cost and an academic focus create obstacles for applying this software in bioprocessing.
Easy to implement software tool
Schaeffer’s team listed a collection of existing tools, but found each one limited in some way, from cost to complexity. So, the scientists used MATLAB to develop an app, GlyKAn AZ, that identifies glycans, is easy to use, and can be customized. In fact, Schaeffer and her colleagues reported that it is even “easy to implement the software tool in biopharmaceutical analytical laboratories.” Plus, this app is freely available on GitHub.
In bioprocessing, advanced analysis also needs to be fast. Even running on a laptop, GlyKAn AZ can analyze the glycans in a sample in just minutes. As the scientists noted however: “This analysis time depends on the complexity of the data set, the analyst’s prior knowledge of the glycan spectra, and the extent of MS/MS fragment annotation desired.”
By starting with MATLAB, a well-known software tool, Schaeffer’s team showed that scientists working on biotherapeutics don’t always need to completely reinvent the wheel behind bioanalysis. Sometimes, more useful applications arise from making more use of existing tools—using that foundation to develop just what is needed by many scientists in the bioprocessing industry.
