• Sequential recommendation is a fundamental task for network applications, and it usually suffers from the item cold start problem due to the insufficiency of user feedbacks. There are currently three kinds of popular approaches which are respectively based on matrix factorization (MF) of collaborative filtering, Markov chain (MC), and recurrent neural network (RNN). Although widely used, they have some limitations. MF based methods could not capture dynamic user's interest. The strong Markov assumption greatly limits the performance of MC based methods. RNN based methods are still in the early stage of incorporating additional information. Based on these basic models, many methods with additional information only validate incorporating one modality in a separate way. In this work, to make the sequential recommendation and deal with the item cold start problem, we propose a Multi-View Recurrent Neural Network (MV-RNN}) model. Given the latent feature, MV-RNN can alleviate the item cold start problem by incorporating visual and textual information. First, At the input of MV-RNN, three different combinations of multi-view features are studied, like concatenation, fusion by addition and fusion by reconstructing the original multi-modal data. MV-RNN applies the recurrent structure to dynamically capture the user's interest. Second, we design a separate structure and a united structure on the hidden state of MV-RNN to explore a more effective way to handle multi-view features. Experiments on two real-world datasets show that MV-RNN can effectively generate the personalized ranking list, tackle the missing modalities problem and significantly alleviate the item cold start problem.
  • We report vibrational sum frequency generation spectra from supported lipid bilayers in which the OH and the CH stretching signals are probed at different salt concentrations. Atomistic simulations show a negligible impact of salt on the OH stretching spectra, indicating the observed SFG intensity changes are due to chi(3) and potential dependent contributions. These are further analyzed in the contact of exact-zero reference states. Further experiments and simulations identify specific hydrogen bonding interactions between interfacial water molecules at the PC head group of the zwitterionic DMPC lipids at 3200 wavenumbers.
  • Sequential recommendation is one of fundamental tasks for Web applications. Previous methods are mostly based on Markov chains with a strong Markov assumption. Recently, recurrent neural networks (RNNs) are getting more and more popular and has demonstrated its effectiveness in many tasks. The last hidden state is usually applied as the sequence's representation to make recommendation. Benefit from the natural characteristics of RNN, the hidden state is a combination of long-term dependency and short-term interest to some degrees. However, the monotonic temporal dependency of RNN impairs the user's short-term interest. Consequently, the hidden state is not sufficient to reflect the user's final interest. In this work, to deal with this problem, we propose a Hierarchical Contextual Attention-based GRU (HCA-GRU) network. The first level of HCA-GRU is conducted on the input. We construct a contextual input by using several recent inputs based on the attention mechanism. This can model the complicated correlations among recent items and strengthen the hidden state. The second level is executed on the hidden state. We fuse the current hidden state and a contextual hidden state built by the attention mechanism, which leads to a more suitable user's overall interest. Experiments on two real-world datasets show that HCA-GRU can effectively generate the personalized ranking list and achieve significant improvement.
  • Semi-empirical quantum mechanical methods traditionally expand the electron density in a minimal, valence-only electron basis set. The minimal-basis approximation causes molecular polarization to be underestimated, and hence intermolecular interaction energies are also underestimated, especially for intermolecular interactions involving charged species. In this work, the third-order self-consistent charge density functional tight-binding method (DFTB3) is augmented with an auxiliary response density using the chemical-potential equalization (CPE) method and an empirical dispersion correction (D3). The parameters in the CPE and D3 models are fitted to high-level CCSD(T) reference interaction energies for a broad range of chemical species, as well as dipole moments calculated at the DFT level; the impact of including polarizabilities of molecules in the parameterization is also considered. Parameters for the elements H, C, N, O and S are presented. The RMSD interaction energy is improved from 6.07 kcal/mol to 1.49 kcal/mol for interactions with one charged specie, whereas the RMSD is improved from 5.60 kcal/mol to 1.73 for a set of 9 salt bridges, compared to uncorrected DFTB3. For large water clusters and complexes that are dominated by dispersion interactions, the already satisfactory performance of the DFTB3-D3 model is retained; polarizabilities of neutral molecules are also notably improved. Overall, the CPE extension of DFTB3-D3 provides a more balanced description of different types of non-covalent interactions than NDDO type of semi-empirical methods (e.g., PM6-D3H4) and PBE-D3 with modest basis sets.
  • Normal mode analysis offers an efficient way of modeling the conformational flexibility of protein structures. Simple models defined by contact topology, known as elastic network models, have been used to model a variety of systems, but the validation is typically limited to individual modes for a single protein. We use anisotropic displacement parameters from crystallography to test the quality of prediction of both the magnitude and directionality of conformational variance. Normal modes from four simple elastic network model potentials and from the CHARMM forcefield are calculated for a data set of 83 diverse, ultrahigh resolution crystal structures. While all five potentials provide good predictions of the magnitude of flexibility, the methods that consider all atoms have a clear edge at prediction of directionality, and the CHARMM potential produces the best agreement. The low-frequency modes from different potentials are similar, but those computed from the CHARMM potential show the greatest difference from the elastic network models. This was illustrated by computing the dynamic correlation matrices from different potentials for a PDZ domain structure. Comparison of normal mode results with anisotropic temperature factors opens the possibility of using ultrahigh resolution crystallographic data as a quantitative measure of molecular flexibility. The comprehensive evaluation demonstrates the costs and benefits of using normal mode potentials of varying complexity. Comparison of the dynamic correlation matrices suggests that a combination of topological and chemical potentials may help identify residues in which chemical forces make large contributions to intramolecular coupling.
  • Simple coarse-grained models, such as the Gaussian Network Model, have been shown to capture some of the features of equilibrium protein dynamics. We extend this model by using atomic contacts to define residue interactions and introducing more than one interaction parameter between residues. We use B-factors from 98 ultra-high resolution X-ray crystal structures to optimize the interaction parameters. The average correlation between GNM fluctuation predictions and the B-factors is 0.64 for the data set, consistent with a previous large-scale study. By separating residue interactions into covalent and noncovalent, we achieve an average correlation of 0.74, and addition of ligands and cofactors further improves the correlation to 0.75. However, further separating the noncovalent interactions into nonpolar, polar, and mixed yields no significant improvement. The addition of simple chemical information results in better prediction quality without increasing the size of the coarse-grained model.