De Novo Models For Raman: Comparison To And Advantages Over Traditional Chemometrics Models

As Raman spectroscopy, with partial least squares (PLS) models, becomes more popular in bioprocesses applications it's important to recognize the challenges this approach presents. Raman spectroscopy is widely used in pharmaceuticals, offering specificity and flexibility in measurement. However, its interpretation requires computational assistance, which faces challenges due to process variations and covariance among chemical species. The MAVERICK de novo model avoids empirical data and addresses issues faced by conventional methods. It utilizes specific Raman reference information and dynamic system characteristics to make real-time adjustments during bioprocess runs. By focusing on major components, it reduces complexity and accounts for system errors, offering direct estimation of sensitivity and selectivity. The de novo model presents a promising solution to the complications encountered in Raman spectroscopy for bioprocess applications, offering improved accuracy and efficiency.