Machine learning methods can be used as robust techniques to provide invaluable information for analyzing biological samples in pharmaceutical industries, such as predicting the concentration of viral particles of interest in biological samples. Here, we utilized both convolutional neural networks and random forests to predict the concentration of the samples containing measles, mumps, rubella, and varicella-zoster viruses (ProQuad®) based on Raman and absorption spectroscopy. We prepared Raman and absorption spectra datasets with known concentration values, then used the Raman and absorption signals individually and together to train RFs and CNNs. We demonstrated that both RFs and CNNs can make predictions with R2 values as high as 95%. We proposed two different networks to jointly use the Raman and absorption spectra, where our results demonstrated that concatenating the Raman and absorption data increases the prediction accuracy compared to using either Raman or absorption spectrum alone. Additionally, we further verified the advantage of using joint Raman-absorption with principal component analysis (PCA). Furthermore, our method can be extended to characterize properties other than concentration, such as the type of viral particles.

We introduce a new method to measure the concentration-dependent diffusion coefficient from a sequence of images of molecules diffusion from a source towards a sink. Generally, approaches measuring the diffusion coefficient, such as Fluorescence Recovery After Photobleaching (FRAP), assume that the diffusion coefficient is constant. Hence, these methods cannot capture the concentration dependence of the diffusion coefficient if present. Other approaches measure the concentration-dependent diffusion coefficient from an instantaneous concentration profile and lose the temporal information. These methods make unrealistic assumptions and lead to 100% error. We introduce a novel image analysis framework that utilizes spatial and temporal information in a sequence of concentration images and numerically solves the general form of Fick’s law using Radial Basis Functions (RBF) to measure the concentration-dependent diffusion coefficient. We term this approach as Concentration Image Diffusimetry (CID). Our method makes no assumptions about the sink and source size. CID is superior to existing methods in estimating spatiotemporal changes and concentration-dependent diffusion. CID also provides a statistical uncertainty quantification on the measurements using bootstrapping, improving the reliability of the diffusion measurement. We assessed CID’s performance using synthetically generated images. Our analysis suggests that CID measures the diffusion coefficient with less than 2% error for most cases. We validated CID with FRAP experimental images and showed that CID agrees with established FRAP algorithms for constant diffusion coefficient. Finally, we demonstrate the application of CID to experimental datasets of a concentration gradient-driven protein diffusion into a tissue replicate.

Deep learning has the potential to revolutionize process analytical technology in the pharmaceutical industry. Here, we used Raman spectroscopy-based deep learning strategies to develop a tool for detecting microbial contamination. We built a Raman dataset for microorganisms that are common contaminants in the pharmaceutical industry for Chinese Hamster Ovary (CHO) cells, which are often used in the production of biologics. Using a convolution neural network (CNN), we classified the different samples comprising individual microbes and microbes mixed with CHO cells with an accuracy of 95-100%. The set of 12 microbes spans across Gram-positive and Gram-negative bacteria as well as fungi. We also created an attention map for different microbes and CHO cells to highlight which segments of the Raman spectra contribute the most to help discriminate between different species. This dataset and algorithm provide a route for implementing Raman spectroscopy for detecting microbial contamination in the pharmaceutical industry.