We propose a new framework for estimating generative models via an adversarial process, in which we simultaneously train two models: a generative model G that captures the data distribution, and a discriminative model D that estimates the probability that a sample came from the training data rather than G. The training procedure for G is to maximize the probability of D making a mistake. This framework corresponds to a minimax two-player game. In the space of arbitrary functions G and D, a unique solution exists, with G recovering the training data distribution and D equal to 1 2 everywhere. In the case where G and D are defined by multilayer perceptrons, the entire system can be trained with backpropagation. There is no need for any Markov chains or unrolled approximate inference networks during either training or generation of samples. Experiments demonstrate the potential of the framework through qualitative and quantitative evaluation of the generated samples.

In this review we intend to provide a relatively comprehensive summary of the work of supramolecular hydrogelators after 2004 and to put emphasis particularly on the applications of supramolecular hydrogels/hydrogelators as molecular biomaterials. After a brief introduction of methods for generating supramolecular hydrogels, we discuss supramolecular hydrogelators on the basis of their categories, such as small organic molecules, coordination complexes, peptides, nucleobases, and saccharides. Following molecular design, we focus on various potential applications of supramolecular hydrogels as molecular biomaterials, classified by their applications in cell cultures, tissue engineering, cell behavior, imaging, and unique applications of hydrogelators. Particularly, we discuss the applications of supramolecular hydrogelators after they form supramolecular assemblies but prior to reaching the critical gelation concentration because this subject is less explored but may hold equally great promise for helping address fundamental questions about the mechanisms or the consequences of the self-assembly of molecules, including low molecular weight ones. Finally, we provide a perspective on supramolecular hydrogelators. We hope that this review will serve as an updated introduction and reference for researchers who are interested in exploring supramolecular hydrogelators as molecular biomaterials for addressing the societal needs at various frontiers.

The combination of nanotechnology and molecular biology has developed into an emerging research area: nanobiotechnology. Magnetic nanoparticles are well-established nanomaterials that offer controlled size, ability to be manipulated externally, and enhancement of contrast in magnetic resonance imaging (MRI). As a result, these nanoparticles could have many applications in biology and medicine, including protein purification, drug delivery, and medical imaging. Because of the potential benefits of multimodal functionality in biomedical applications, researchers would like to design and fabricate multifunctional magnetic nanoparticles. Currently, there are two strategies to fabricate magnetic nanoparticle-based multifunctional nanostructures. The first, molecular functionalization, involves attaching antibodies, proteins, and dyes to the magnetic nanoparticles. The other method integrates the magnetic nanoparticles with other functional nanocomponents, such as quantum dots (QDs) or metallic nanoparticles. Because they can exhibit several features synergistically and deliver more than one function simultaneously, such multifunctional magnetic nanoparticles could have unique advantages in biomedical applications. In this Account, we review examples of the design and biomedical application of multifunctional magnetic nanoparticles. After their conjugation with proper ligands, antibodies, or proteins, the biofunctional magnetic nanoparticles exhibit highly selective binding. These results indicate that such nanoparticles could be applied to biological medical problems such as protein purification, bacterial detection, and toxin decorporation. The hybrid nanostructures, which combine magnetic nanoparticles with other nanocomponents, exhibit paramagnetism alongside features such as fluorescence or enhanced optical contrast. Such structures could provide a platform for enhanced medical imaging and controlled drug delivery. We expect that the combination of unique structural characteristics and integrated functions of multicomponent magnetic nanoparticles will attract increasing research interest and could lead to new opportunities in nanomedicine.