Search

Modeling the impact of amino acid mutations on protein-protein interaction plays a crucial role in protein engineering and drug design. In this study, we develop GeoPPI, a novel structure-based deep-learning framework to predict the change of binding affinity upon mutations. Based on the three-dimensional structure of a protein, GeoPPI first learns a geometric representation that encodes topology features of the protein structure via a self-supervised learning scheme. These representations are then used as features for training gradient-boosting trees to predict the changes of protein-protein binding affinity upon mutations. We find that GeoPPI is able to learn meaningful features that characterize interactions between atoms in protein structures. In addition, through extensive experiments, we show that GeoPPI achieves new state-of-the-art performance in predicting the binding affinity changes upon both single- and multi-point mutations on six benchmark datasets. Moreover, we show that GeoPPI can accurately estimate the difference of binding affinities between a few recently identified SARS-CoV-2 antibodies and the receptor-binding domain (RBD) of the S protein. These results demonstrate the potential of GeoPPI as a powerful and useful computational tool in protein design and engineering. Our code and datasets are available at: https://github.com/Liuxg16/GeoPPI., Author summary Estimating the binding affinities of protein-protein interactions (PPIs) is crucial to understand protein function and design new functional proteins. Since the experimental measurement in wet-labs is labor-intensive and time-consuming, fast and accurate in silico approaches have received much attention. Although considerable efforts have been made in this direction, predicting the effects of mutations on the protein-protein binding affinity is still a challenging research problem. In this work, we introduce GeoPPI, a novel computational approach that uses deep geometric representations of protein complexes to predict the effects of mutations on the binding affinity. The geometric representations are first learned via a self-supervised learning scheme and then integrated with gradient-boosting trees to accomplish the prediction. We find that the learned representations encode meaningful patterns underlying the interactions between atoms in protein structures. Also, extensive tests on major benchmark datasets show that GeoPPI has made an important improvement over the existing methods in predicting the effects of mutations on the binding affinity.

Translation elongation is a crucial phase during protein biosynthesis. In this study, we develop a novel deep reinforcement learning based framework, named Riboexp, to model the uneven distribution of ribosomes on mRNA transcripts during translation elongation. In particular, Riboexp employs a policy network to perform a context-dependent feature selection for the prediction of ribosome density. Our extensive tests demonstrate that Riboexp can outperform other state-of-the-art methods in predicting ribosome density by up to 5.9% in per-gene Pearson correlation coefficient. Enabled by the reinforcement learning strategy, Riboexp can identify more relevant features for the prediction task than other commonly used attribution methods in deep learning. In-depth analyses also reveal the meaningful biological insights generated by the Riboexp framework. Moreover, the application of Riboexp in codon optimization results in an increase of protein production by around 31% over the previous state-of-the-art method that models ribosome density. These results have established Riboexp as a powerful and useful computational tool in the studies of translation dynamics and protein synthesis.

Chemical retrosynthesis has been a crucial and challenging task in organic chemistry for several decades. In early years, retrosynthesis is accomplished by the disconnection approach which is labor-intensive and requires expert knowledge. Afterward, rule-based methods have dominated in retrosynthesis for years. In this study, we revisit the disconnection approach by leveraging deep learning (DL) to boost its performance and increase the explainability of DL. Concretely, we propose a novel graph-based deep-learning framework, named DeRetro, to predict the set of reactants for a target product by executing the process of disconnection and reactant generation orderly. Experimental results report that DeRetro achieves new state-of-the-art performance in predicting the reactants. In-depth analyses also demonstrate that even without the reaction type as input, DeRetro retains its retrosynthesis performance while other methods show a significant decrease, resulting in a large margin of 19% between DeRetro and previous state-of-the-art rule-based method. These results have established DeRetro as a powerful and useful computational tool in solving the challenging problem of retrosynthetic analysis.

Translation elongation plays a crucial role in multiple aspects of protein biogenesis. In this study, we develop a novel deep reinforcement learning based framework, named RiboRL, to model the distributions of ribosomes on transcripts. In particular, RiboRL employs a policy network (PolicyNet) to perform a context-dependent feature selection to facilitate the prediction of ribosome density. Extensive tests demonstrate that RiboRL can outperform other state-of-the-art methods in predicting ribosome densities. We also show that the reinforcement learning based strategy can generate more informative features for the prediction task when compared to other commonly used attribution methods in deep learning. Moreover, the in-depth analyses and a case study also indicate the potential applications of the RiboRL framework in generating meaningful biological insights regarding translation elongation dynamics. These results have established RiboRL as a useful computational tool to facilitate the studies of the underlying mechanisms of translational regulation.

We propose Object-oriented Neural Programming (OONP), a framework for semantically parsing documents in specific domains. Basically, OONP reads a document and parses it into a predesigned object-oriented data structure (referred to as ontology in this paper) that reflects the domain-specific semantics of the document. An OONP parser models semantic parsing as a decision process: a neural net-based Reader sequentially goes through the document, and during the process it builds and updates an intermediate ontology to summarize its partial understanding of the text it covers. OONP supports a rich family of operations (both symbolic and differentiable) for composing the ontology, and a big variety of forms (both symbolic and differentiable) for representing the state and the document. An OONP parser can be trained with supervision of different forms and strength, including supervised learning (SL) , reinforcement learning (RL) and hybrid of the two. Our experiments on both synthetic and real-world document parsing tasks have shown that OONP can learn to handle fairly complicated ontology with training data of modest sizes., Comment: accepted by ACL 2018

In this paper, improved genetic algorithm (GA) is used to search Chang'e-3 optimal landing preparative orbit perilune. Chang'e-3 lander as an object, the initial and final state of the lander and the kinetic equations in the lunar gravitational field are determined by reversely reasoning from the predetermined landing site in the main reduction stage. The lander motion state is non-linear and continuous, so this method takes advantage of genetic algorithms and discretized track to achieve the optimal location of perilune, with the help of Kepler's third law and the law of inertia. Simulation results show that the improved genetic algorithm has the better performance in convergence speed and can be applied to search for the global optimal location of perilune.

This paper proposes a method to search safe landing site using dual-threshold image segmentation algorithm for Chang'e-3 when the digital elevation map is obtained. Aiming at meeting the need of security, the probe ought to choose a flat ground to land at. Thus an improved dual-threshold image segmentation algorithm is proposed to deal with the shadowed and light split regions of obstacles respectively. At the same time, genetic algorithm is used to optimize its parameters, and the image sampling is also used to improve the efficiency of the algorithm. In order to identify candidate safe landing areas, a search method based on the maximum blank squares is applied in the processed image. Experimental results show that the proposed method has a strong ability of recognizing obstacles and searching flat ground, and is conductive to real-time space exploration mission.