Reporter：Professor Huang Shengyou of Huazhong University of Bet365 login and Technology
Title: Bet365 login prediction: from molecular docking toAIIntegrated modeling
Time: November 26 (Tuesday) 10 am
Location：Computer Building313Conference Room
Report Summary:
Bet365 login is the executor of life activities.“Structure determines function”，Therefore，Determining protein interactions and their complex structures is critical to understanding life activities and drug development。Due to its clear physical principles and extremely low computational cost，Molecular docking has always been the most important fast calculation method in Bet365 login prediction，However，Due to limitations in molecular flexibility and energy scoring functions，The accuracy has reached the bottleneck。In recent years，Artificial intelligence model such asAlphaFold2Achieved great success in protein monomer structure prediction，But due to insufficient samples，The accuracy of Bet365 login prediction still needs to be improved。Therefore，New methods are urgently needed to break through the bottleneck of existing Bet365 login prediction。In recent years，Cryo-EM has developed into the most important experimental method for determining the structure of biological macromolecules，The information contained in its experimental data exactly meets the needs of Bet365 login prediction。In this report，I will first introduce the Bet365 login prediction method based on molecular docking，Then we will focus on reporting on our mining of complex structure information from cryo-electron microscopy experimental density maps based on artificial intelligence，And integrate the mined experimental information to conduct research on Bet365 login modeling。
About the speaker:
Huang Shengyou，Professor of Huazhong University of Bet365 login and Technology、Ph.D. Supervisor，National leading talents，Winner of 100 outstanding doctoral dissertations nationwide。Long-term research on protein interaction calculations and complex structure prediction，Total papers published120Remaining articles, recent5Corresponding author in yearNature Biotechnology、Nature Machine Intelligence、Nature CommunicationsPublish papers in other magazines40Remaining articles, developedHDOCKProtein molecular docking algorithm in international Bet365 login predictionCAPRIRanked first in competitions many times。Host the National Natural Bet365 login Foundation of China key project、International cooperation projects, etc.。Current Chinese Biological Informatics Society(raise)"Biomolecule Structure Prediction and Simulation Professional Committee”Secretary-General.
Research group website:http://huanglab.phys.hust.edu.cn/。
Reporter: Researcher Zhu Shanfeng of Fudan Bet365 login
Title：GORetriever: reranking Bet365 login-description-based GO candidates by literature-driven deep information retrieval for Bet365 login function annotation
Time：11month26Sunday (Tuesday) morning10Points
Location：Computer Building313Conference Room
Report Summary:
The vast majority of proteins still lack experimentally validated functional annotations, which highlights the importance of developing high-performance automated Bet365 login function prediction/annotation (AFP) methods. While existing approaches focus on Bet365 login sequences, networks, and structural data, textual information related to proteins has been overlooked. However, roughly 82% of SwissProt proteins already possess literature information that experts have annotated. To efficiently and effectively use literature information, we present GORetriever, a two-stage deep information retrieval-based method for AFP. Given a target Bet365 login, in the first stage, candidate Gene Ontology (GO) terms are retrieved by using annotated proteins with similar descriptions. In the second stage, the GO terms are reranked based on semantic matching between the GO definitions and textual information (literature and Bet365 login description) of the target Bet365 login. Extensive experiments over benchmark datasets demonstrate the remarkable effectiveness of GORetriever in enhancing the AFP performance. Note that GORetriever is the key component of GOCurator, which has achieved the first place in the latest critical assessment of Bet365 login function annotation (CAFA5: over 1,600 teams participated), held in 2023–24.
About the speaker:
Zhu Shanfeng，Researcher at Fudan University Institute of Brain-inspired Intelligence Bet365 login and Technology，Doctoral Supervisor。UniProt Member of the International Scientific Advisory Board, PLoS Computational BiologyEditor, Oxford Bet365 login PressBioinformatics Advances Deputy Editor。Has hosted five National Natural Bet365 login Foundation projects，And multiple domestic and foreign enterprise R&D projects。The main research direction is artificial intelligence and biomedical big data mining，Especially biomedical text mining、Protein function prediction、Metagenomic、Smart medical care, etc.。Relevant papers listed as first or corresponding author in biological information、Artificial Intelligence、Published in data mining and other related flagship international conferences and journals，such as NeurIPS, KDD, ISMB, IJCAI, ACL, Nature Communications、Genome Biology, Bioinformatics, Nucleic Acids Researchetc.2014Year-2022 year BioASQ Won first place eight times in an international competition for automatic annotation of large-scale biomedical texts。2017、2020and2023 Participate in each year CAFA3、CAFA4andCAFA5 Won first place in the international competition for automatic annotation of large-scale Bet365 login functions。2019Year-2021yearCAMI IIInternational competition for large-scale metagenomic data analysis ranked first overall in contig binning algorithm。
