Prediction of Human Induced Pluripotent Stem Cell Cardiac Differentiation Outcome by Multifactorial Process Modeling

Abstract

Human cardiomyocytes (CMs) have potential for use in therapeutic cell therapy and high-throughput drug screening. Because of the inability to expand adult CMs, their large-scale production from human pluripotent stem cells (hPSC) has been suggested. Significant improvements have been made in understanding directed differentiation processes of CMs from hPSCs and their suspension culture-based production at chemically defined conditions. However, optimization experiments are costly, timeconsuming, and highly variable, leading to challenges in developing reliable and consistent protocols for the generation of large CM numbers at high purity. This study examined the ability of data-driven modeling with machine learning for identifying key experimental conditions and predicting final CM content using data collected during hPSC-cardiac differentiation in advanced stirred tank bioreactors (STBRs). Through feature selection, we identified process conditions, features, and patterns that are the most influential on and predictive of the CM content at the process endpoint, on differentiation day 10 (dd10). Process-related features were extracted from experimental data collected from 58 differentiation experiments by feature engineering. These features included data continuously collected online by the bioreactor system, such as dissolved oxygen concentration and pH patterns, as well as offline determined data, including the cell density, cell aggregate size, and nutrient concentrations. The selected features were used as inputs to construct models to classify the resulting CM content as being "sufficient" or "insufficient" regarding pre-defined thresholds. The models built using random forests and Gaussian process modeling predicted insufficient CM content for a differentiation process with 90% accuracy and precision on dd7 of the protocol and with 85% accuracy and 82% precision at a substantially earlier stage: dd5. These models provide insight into potential key factors affecting hPSC cardiac differentiation to aid in selecting future experimental conditions and can predict the final CM content at earlier

Cellular and Engineered Organoids for Cardiovascular Models

“…However, the main difference between the resulting constructs after spontaneous cardiogenesis and cellular assemblage lies in the patterning or spatially organized features. 99 COs are generally used as broad term for precardiac structures and feature-specific organoids, which, in part, constitute the cell or cardiac tissue-like assembly in a miniature form. Terminologies, such as gastruloids, cardiods, heart-forming organoids, and heart organoid, are confusing due to their striking differences in the features exhibited by different CO models.…”

“…However, the main difference between the resulting constructs after spontaneous cardiogenesis and cellular assemblage lies in the patterning or spatially organized features. 99…”

Induced pluripotency in the context of stem cell expansion bioprocess development, optimization, and manufacturing: a roadmap to the clinic

“…There exist unique opportunities to apply concepts in machine learning and computational modeling to advance stem-cell bioprocess development and clinical manufacturing. Several works within basic research and clinical biomanufacturing spheres have been published with this mindset, though the field of iPSC biomanufacturing has yet to rigorously explore these opportunities 82 – 84 . Robust iPSC expansion and differentiation processes are closed-loop systems, and the incorporation of predictive models and real-time intervention strategies has the potential to greatly enhance predictability and online process control over cell health and phenotype, resulting in safer and more effective iPSC-derived cell therapeutics.…”

Classification of cardiac differentiation outcome, percentage of cardiomyocytes on day 10 of differentiation, for hydrogel‐encapsulated hiPSCs

“…Naïve Bayes, support vector machines, and Knearest neighbors were used to classify the diseased and abnormal CMs. [17,18] Williams et al [19] predicted the outcome for producing CMs, which were differentiated from hiPSCs in a bioreactor, using different models, such as RF and GP modeling. All these studies demonstrated the promising results of using ML techniques to classify, predict, and select features for constructing accurate predictive models for 3D CM production systems and their potential to be used in scale-up analysis.…”