Physiologically, a hallmark of biological hydrogels is their ability to selectively trap diffusing molecules and particles. And indeed, there is now increasing interest in using selective hydrogel barriers for applications in biomedicine and medical engineering. However, when employing synthetic polymers to create hydrogels with selective permeabilities, controlling the type and strength of the ensuing filtration process is difficult. Here, we generate hybrid gels with adjustable selectivity profiles by mixing a series of (bio‐)macromolecules with agarose. Depending on the type and concentration of the incorporated macromolecules, those hybrid gels achieve a selective retardation of the diffusive translocation of either positively or negatively charged dextrans at both, acidic and neutral pH. Furthermore, we demonstrate three strategies that provide hydrogels with sequestered patches of both, cationic and anionic binding sites, thus creating symmetric charge (i.e., electrostatic bandpass) filters which still allow neutral molecules to pass. Moreover, such agarose matrices offer a high level of versatility as their permeability profiles can be tailored at will by integrating macromolecules with desired physico‐chemical properties. Thus, those agarose‐based hybrid gels may serve as a powerful platform to engineer adjustable and versatile materials for a broad range of future applications in the field of biomedical engineering.

Mucus is a complex hydrogel biomaterial whose composition is regulated meticulously to ensure that its important function as a selective barrier is maintained. As part of this function, mucus regulates the uptake of molecules from the gastrointestinal lumen into the body. Yet, those hydrogels are continuously challenged with environmental pollutants such as black carbon nanoparticles (NPs), and there is growing evidence that these contaminants can compromise the functionality of mucus. Here, we assess the impact of black carbon NP contaminations on the selective permeability properties of mucin hydrogels. For this purpose, we identified two physiologically relevant black carbon concentrations and used those NP concentrations to perform molecular penetration studies with pristine and contaminated mucin hydrogels. We found that the presence of black carbon NPs enhances both the partitioning of anionic molecules into mucin hydrogels and the translocation of cationic molecules across those barriers. Moreover, we found that this permeability modulating effect is asymmetric with respect to charge; i.e., the penetration and translocation behavior of cationic molecules is affected more strongly than that of anionic ones. To rationalize those findings, we propose that black carbon NPs are well integrated into the mucin glycoprotein network, thus masking more anionic binding sites on mucins than creating cationic ones. Our results underscore the high value of suitable in vitro models when trying to decipher the nanoscopic effects by which physiologically relevant contaminants can influence molecular transport phenomena across mucosal barriers.

With big datasets and highly efficient algorithms becoming increasingly available for many problem sets, rapid advancements and recent breakthroughs achieved in the field of machine learning encourage more and more scientific fields to make use of such a computational data analysis. Still, for many research problems, the amount of data available for training a machine learning (ML) model is very limited. An important strategy to combat the problems arising from data sparsity is feature elimination—a method that aims at reducing the dimensionality of an input feature space. Most such strategies exclusively focus on analyzing pairwise correlations, or they eliminate features based on their relation to a selected output label or by optimizing performance measures of a certain ML model. However, those strategies do not necessarily remove redundant information from datasets and cannot be applied to certain situations, e.g., to unsupervised learning models. Neither of these limitations applies to the network-based, correlation-driven redundancy elimination (NETCORE) algorithm introduced here, where the size of a feature vector is reduced by considering both redundancy and elimination efficiency. The NETCORE algorithm is model-independent, does not require an output label, and is applicable to all kinds of correlation topographies within a dataset. Thus, this algorithm has the potential to be a highly beneficial preprocessing tool for various machine learning pipelines.