A New Fingerprint and Graph Hybrid Neural Network for Predicting Molecular Properties.

Machine learning plays a role in accelerating drug discovery, and the design of effective machine learning models is crucial for accurately predicting molecular properties. Characterizing molecules typically involves the use of molecular fingerprints and molecular graphs. These are input into a multilayer perceptron (MLP) and variants of graph neural networks, such as graph attention networks (GATs). Due to the diverse types and large dimension of fingerprints, models may contain many features that are relatively irrelevant or redundant; meanwhile, although the GAT excels in handling heterogeneous graph tasks, it lacks the ability to extract collaborative information from neighboring nodes, which is crucial in scenarios where it cannot capture the joint influence of adjacent groups on atoms. To overcome these challenges, we introduce a hybrid model, combining improved GAT and MLP. In GAT, the recurrent neural network is employed to capture collaborative information. To address the dimensionality issue, we propose a feature selection algorithm, which is based on the principle of maximizing relevance while minimizing redundancy. Through experiments on 13 public data sets and 14 breast cell lines, our model demonstrates superior performance compared to state-of-the-art deep learning and traditional machine learning algorithms. Additionally, a series of ablation experiments were conducted to demonstrate the advantages of our improved version, as well as its antinoise capability and interpretability. These results indicate that our model holds promising prospects for practical applications.

Fullerene-intercalated graphene nanocontainers for gas storage and sustained release.

Molecular dynamics simulations are performed to investigate the storage capacity and sustained release of nitrogen (N2) in the graphene-based nanocontainers. Sandwiched graphene-fullerene composites (SGFC) composed of two parallel graphene sheets and intercalated fullerenes are constructed. The simulation results show that the mass density of N2 at the first layer is extremely high, due to the strong adsorption ability of graphene sheets. And N2 molecules at this adsorbed layer are thermodynamically stable. Furthermore, we analyze the storage efficiency of SGFC. In general, the gravimetric and volumetric capacities decrease with the increasing number of intercalated fullerenes. On the contrary, the stability of SGFC is enhanced by more intercalated fullerenes. We therefore make a compromise and propose that 1 fullerene per 5 nm2 graphene to build a SGFC, which is much effective to storage N2. We also verify the reversibility that N2 can sustainably release from the SGFC. Our results may provide insights into the design of graphene-based nanocomposites for gas storage and sustained release with excellent structural stability and high storage capacity. Graphical abstract.