Chromatography is a key unit operation in the biopharmaceutical manufacturing process used for protein purification and polishing. Design and optimization of these processes are resource-intensive resulting from the complex combinatorial design space. This constraint combined with the wide diversity in therapeutic formats and increased pressure for timely delivery to the market necessitates an efficient, fast, robust and generalized framework for process design and optimization. Here we present Gaussian processes as a potent machine learning methodology for predictive modeling in the context of QSAR modeling used for resin and solvent condition selection. We highlight the on-par predictive power of Gaussian Processes with other reported machine learning algorithms. Furthermore, we demonstrate additional properties of Gaussian processes such as its ability to provide confidence estimates for its prediction that makes it suitable for model-assisted optimization. Finally, we demonstrate the possibility to derive feature importances from Gaussian processes, making these models as interpretable as ensembled tree-based methods such as random forests.

Study of the concentration effect by rheological measurements for ultrasound and heat-treated native soy protein isolate

SPI/HG composite microgels with tunable structure and rheological properties for semi-solid foods.

Enhanced gelation of myofibrillar protein in frozen Litopenaeus vannamei: through the physical filling of natural soy protein isolate and cross-linking of denatured soy protein isolate.

A Self-Cleavable ULP1-SUMO Fusion Strategy for Enhanced Soluble Protein Expression and One-Step Purification in Escherichia coli.

Comparative characterization of protein isolates from treated and un-treatedRed kidney bean (Phaseolus vulgaris) and horsegram (Macrotyloma uniflorum) with commercial pea protein (Pisum Sativum): Nutritional, functional, in vitro digestibility and structural perspectives.

Yellow pea germination impact on protein isolation - protein digestibility and quality

Alkaline-pH-shifted mung bean protein isolate for spray-dried probiotic encapsulation: Insights into molecular cell wall interactions and oxidative protection.

In vitro protein digestibility and protein quality assessment of novel brewers' spent grain protein isolates

Recombinant expression and purification of TANB77-derived pilin-like proteins 935 and 938.

Molecularly imprinted technology （MIT） represents an advanced synthetic strategy that emulates biological recognition mechanisms， such as antigen-antibody or enzyme-substrate interactions， by creating three-dimensional cavity-like structures through the directional assembly of functional monomers around a template molecule. This process generates spatial and functional complementarity， enabling highly selective recognition of target species. Molecularly imprinted polymers （MIPs）， often described as "synthetic antibodies"， overcome the intrinsic limitations of natural biomolecules by offering superior selectivity， robustness， cost-effectiveness， and structural tunability. These features position MIPs as promising alternatives to natural antibodies in targeted sensing and drug delivery， with broad applications across biomedical， environmental， and pharmaceutical domains， including pollutant detection， and food safety monitoring. Despite substantial progress， key challenges remain， such as uneven imprinting layers， template residue， and limited aqueous compatibility in macromolecular imprinting. Furthermore， issues of industrial scalability， unclear recognition mechanisms， and insufficient integration with emerging fields such as microfluidics and artificial intelligence have hindered large-scale translation. In recent years， our research team has systematically advanced MIT through a tri-dimensional strategy encompassing high-throughput monomer screening， mechanistic elucidation of molecular recognition， and directional assembly of functional units. By establishing a standardized monomer library and integrating molecular dynamics simulations， we achieved precise material design under complex conditions. Through process optimization and material innovation， we developed a highly efficient solid-phase surface imprinting method that enables the fabrication of smart MIPs with stimuli-responsive properties （e.g.， temperature and pH）. These MIPs exhibit markedly enhanced binding affinity， with equilibrium dissociation constant （KD） reaching 10-12 mol/L， over four orders of magnitude higher than those of non-imprinted polymers （NIPs）. Building on these advances， we established cross-disciplinary application platforms， including affinity-based protein separation and purification systems capable of efficient dual-enzyme cascade immobilization and inactivated enzyme renaturation. In the biomedical domain， we developed ultrasensitive biosensing methods achieving picogram-level detection of heart failure biomarkers and single-digit （≈5 cells/mL） detection of cancer cells in whole blood， extending these methods toward integrated tumor theranostics and microbial community regulation. This paper comprehensively summarizes our team's recent innovations in the rational design， functionalized fabrication， and cross-disciplinary applications of MIPs， spanning biosensing， biocatalysis， and biomedical diagnostics/therapeutics， while contextualizing these within the latest global advances in biomedicine and catalysis. Looking forward， we identify three strategic research frontiers for next-generation MIT. （i） Smart responsive material systems： design MIPs capable of multi-stimuli responsiveness （e.g.， magnetic， photothermal， and pH cues） to enable programmable drug release， real-time signal monitoring， and dynamic feedback regulation. （ii） Quantitative modeling of dynamic recognition： establish multi-scale theoretical frameworks to elucidate coupling between cavity flexibility and target conformational dynamics， guiding structure optimization and function-oriented design of adaptive MIPs. （iii） Integrated intelligent theranostic platforms： integrate microfluidics and biomimetic recognition modules into closed-loop systems capable of biomarker detection， targeted delivery， and real-time therapeutic feedback， bridging the gap between in vitro sensing and in vivo precision intervention. Synergistic advancement along these trajectories will empower MIT to transcend its role as a "static recognition material" and evolve into an intelligent， adaptive， and systematic biomedical platform. Such evolution will accelerate the translation of MIT innovations from laboratory to clinic and industry， propelling progress in personalized medicine， point-of-care diagnostics， and synthetic biology， and yielding profound scientific and societal impact.

Effect of culture media and fermentation process on the refolding and purification of rh-GCSF.

Covalent modification of whey protein isolate by rosmarinic Acid: Impacts on physicochemical and functional properties

Impact of Whey Protein on the Physico-Mechanical Properties of Alginate Dialdehyde Scaffolds and Cell Adhesion for Cultivated Meat Applications.

Optimizing wet isolation conditions and characterizing functionality of protein isolates from air-classified starch-rich pulse flours

Enhancing the functional properties of egg yolk through post-treatments following short-term fermentation for application in powdered oils.

Development of chitosan-based smart active films incorporating soy protein isolate-derived metal nanoenzymes and purple carrot extract for blueberry preservation and beef freshness monitoring.

Gel properties of whole soybean curd reinforced by soybean protein isolate amyloid fibrils: Formation mechanism and texture enhancement.

Fabrication of pea protein isolate-folic acid conjugates via free radical grafting: Formation mechanism, stability, and curcumin delivery potential evaluation.

Effects of metal phenolic networks (MPNs) on the structural, IgE binding capacity and functional properties of whey protein isolates: Insights gained from polyphenols constructing MPN