Search for: All records

Recent advances and achievements of artificial intelligence (AI) as well as deep and graph learning models have established their usefulness in biomedical applications, especially in drug–drug interactions (DDIs). DDIs refer to a change in the effect of one drug to the presence of another drug in the human body, which plays an essential role in drug discovery and clinical research. DDIs prediction through traditional clinical trials and experiments is an expensive and time-consuming process. To correctly apply the advanced AI and deep learning, the developer and user meet various challenges such as the availability and encoding of data resources, and the design of computational methods. This review summarizes chemical structure based, network based, natural language processing based and hybrid methods, providing an updated and accessible guide to the broad researchers and development community with different domain knowledge. We introduce widely used molecular representation and describe the theoretical frameworks of graph neural network models for representing molecular structures. We present the advantages and disadvantages of deep and graph learning methods by performing comparative experiments. We discuss the potential technical challenges and highlight future directions of deep and graph learning models for accelerating DDIs prediction.

 

The multicopper oxidase enzyme laccase holds great potential to be used for biological lignin valorization alongside a biocompatible ionic liquid (IL). However, the IL concentrations required for biomass pretreatment severely inhibit laccase activity. Due to their ability to function in extreme conditions, many thermophilic enzymes have found use in industrial applications. The thermophilic fungal laccase from Myceliophthora thermophila was found to retain high levels of activity in the IL [C 2 C 1 Im][EtSO 4 ], making it a desirable biocatalyst to be used for lignin valorization. In contrast to [C 2 C 1 Im][EtSO 4 ], the biocompatibility of [C 2 C 1 Im][OAC] with the laccase was markedly lower. Severe inhibition of laccase activity was observed in 15% [C 2 C 1 Im][OAc]. In this study, the enzyme surface charges were modified via acetylation, succinylation, cationization, or neutralization. However, these modifications did not show significant improvement in laccase activity or stability in [C 2 C 1 Im][OAc]. Docking simulations show that the IL docks close to the T1 catalytic copper, likely interfering with substrate binding. Although additional docking locations for [OAc] - are observed after making enzyme modifications, it does not appear that these locations play a role in the inhibition of enzyme activity. The results of this study could guide future enzyme engineering efforts by showing that the inhibition mechanism of [C 2 C 1 Im][OAc] toward M. thermophila laccase is likely not dependent upon the IL interacting with the enzyme surface. 

Inversion symmetry breaking could lead to the creation of a Rashba–Dresselhauls magnetic field, which plays the key role in spintronic devices. In this work, we propose and develop a composition gradient engineering approach that breaks inversion symmetry into inorganic halide perovskites with strong spin–orbit coupling. We synthesize epitaxial CsPbBr x Cl (3− x ) with Br/Cl composition gradient by a two-step chemical vapor deposition approach. Through optoelectronic measurements, we show the presence of circular photogalvanic effects (CPGEs), evidencing a Rashba-like spin polarized band structure. By spatially resolved photoluminescence spectra, we find that the observed CPGE is likely a cumulative result of inversion symmetry-broken interfaces featured by abrupt and stepwise composition gradient between the pristine and separated daughter phases. Our work suggests an avenue in engineering the spintronic property of halide perovskites for information processing.