The Ning lab is conducting research on Artificial Intelligence (particularly, generative AI), Machine Learning, and Big Data Analytics with applications for emerging critical problems in drug discovery, Biomedicine, Health Informatics, and e-Commerce. Please check the Research page for more information.
News
- Ning Lab is recruiting MS and PhD students and Postdocs (positions available immediately). Please check the Openings Available at Ning Lab 2025 document for more information.
- (09/29/2024) We have postdoc positions available immediately. Please contact PI Ning (ning dot 104 at osu dot edu) directly if you are interested!
- (09/29/2024) We have openings for Ph.D. students for Spring/Autumn 2025 and MS students on project/thesis track.
- (08/18/2024) See how we can use offline-online reinforcement learning to help chemists make real molecules — a new publication on AI4Science!
- (07/27/2024) We presented our work “eCeLLM: Generalizing large language models for e-commerce from large-scale, high-quality instruction data” at the International Conference on Machine Learning (ICML) in Vienna.
- We constructed ECInstruct, the first open-sourced, large-scale, and high-quality benchmark instruction dataset for e-commerce.
- Leveraging ECInstruct, we developed eCeLLM (pronounce: e-sell’em), a series of e-commerce LLMs, by instruction-tuning general-purpose LLMs.
- Data and models are available at HuggingFace, and have been used in the 2024 KDDcup Multi-task Online Shopping Challenge for LLMs.
- (07/27/2024) Large-Language Models can do Chemistry! Check our work on LlaSMol: Advancing large language models for chemistry with a large-scale, comprehensive, high-quality instruction tuning dataset accepted by the Conference on Language Modeling (COLM).
- We constructed SMolInstruct, a large-scale, comprehensive, and high-quality dataset for instruction tuning. It contains 14 selected chemistry tasks and over three million samples.
- Using SMolInstruct, we fine-tuned a set of open-source LLMs named as LlaSMol.
- Data and models are available at HuggingFace.
- (07/24/2024) We use AI/ML to predict molecule properties via auxiliary learning and task-specific adaptation.
- (06/12/2024) NingLab has been awarded a Sanofi iDEA-TECH Award. More exciting news to come!
- (05/13/2024) Precision medicine requires a customized treatment for each individual patient. We developed AI/ML methods to select anti-cancer drugs for different cancer patients, improving precision medicine treatment for cancer patients. See our new papers in npj Precision Oncology and Journal of Chemical Information and Modeling.
- (04/10/204) Dr. Ning is featured as one of the three spotlight speakers in 2024 OSU Research and Innovation Showcase, a signature event of OSU’s Enterprise for Research, Innovation and Knowledge (ERIK).
- (03/06/2024) Fun beyond research: Photograph “warmth” from NingLab selected to the OSU Health Sciences Art Show, and the OSUWC James Art Gallery “Medicine and The Arts” show
- (05/03/2023) Check out our work on G2Retro as a two-step graph generative models for retrosynthesis prediction, which has been published in Communications Chemistry. The code and webportal are available.
- Here is a nice report from WOSU on our work. Here is the OSU news.
- G2Retro is figured in Nature’s Communications Chemistry special collection: Celebrating Women in Chemistry (well, the first author Ziqi Chen and the contact author Dr. Xia Ning are women in Computer Science).
- (02/24/2023) Check out our work on Binding peptide generation for MHC class I proteins with deep reinforcement learning, which has been published in Bioinformatics.
- This paper is among the top 10% most-read manuscripts in Bioinformatics that were published in 2023.
- (02/03/2023) Check out our work on Understanding comorbidities and health disparities related to COVID-19: a comprehensive study of 776 936 cases and 1 362 545 controls in the state of Indiana, USA, which has been published in JAMIA Open.
- (08/10/2022) Check out our work on MHC Class I antigen processing predictions using representation learning, which has been accepted by Cell Reports Methods.
- (03/18/2022) Check out our work on a knowledge graph from clinical trials (CTKG, Scientific Reports), the dataset and the webportal. CTKG can be used to identify similar clinical trials, similar drugs tested in similar clinical trials, and many other retrieval tasks.
- (03/08/2022) Check out our work on transfer learning and representation learning to predict compound activities and the code.
- (02/01/2022) Check out our work on the next-basket recommendation model M2 using preferences, popularities, and transitions (TKDE) and the code.
- (11/08/2021) Check out our work on molecule optimization and generation using deep learning (preprint, Nature Machine Intelligence) and the code. The supplementary materials have a lot of additional discussions and analyses.
- (05/17/2021) Check out our work on peptide-MHC Class I binding prediction.
- (12/02/2020) Check out our work on cognitive biomarker prioritization in Alzheimer’s Disease.
- (11/19/2020) Our early paper has won the 10-years-highest-impact award at the prestigious International Conference of Data Mining (ICDM). See the news at TDAI.
- (08/27/2020) Check out our work on understanding illness associated with breast implants from social media data.
- (08/12/2020) Check out our work on predicting the potential use of existing FDA approved drugs for treating COVID-19.
- (08/08/2020) Check out our recent manuscript on search recommendations for clinical decision-making support with applications in the emergency room.
- (07/21/2020) Ning will teach CSE 5523 Machine Learning and Statistical Pattern Recognition in Autumn, 2020.
- (07/21/2020) Ning will co-teach BMI 8050.01 – 10 Special Topics in BMI: Natural Language Processing in Biomedical Research in Autumn, 2020.
- (07/03/2020) Check out our most recent manuscript on sequential recommendation using light-weighted, hybrid associations models, together with a comprehensive study of the state-of-the-art sequential recommendation methods.
- (06/02/2020) In collaboration with Amazon Web Services, we announced the open-source biological knowledge graph to fight COVID-19 (also available here at CCTS)