JIANAN NIE

ad0e4q@r.postjobfree.com Virginia Tech https://jianannie.github.io/

RESEARCH INTEREST

AI for Science (Chemistry), AI for Healthcare, Trustworthy NLP, LLM Security EDUCATION

Southern University of Science and Technology (SUSTech) Shenzhen, China B. S. in Chemical Science (GPA: 3.78/4.0) Sept. 2018 – Jun. 2022 Core Courses: Linear Algebra, Calculus I, Calculus II, Introduction to Computer Programming, Java, Quantum Chemistry, Material Characterization, Physics, Biology, Organic Chemistry, Inorganic Chemistry, Physical Chemistry, Analytical Chemistry, Inorganic Chemistry, Coordination Chemistry, Organometallics, Polymer Chemistry, Element- Organic Chemistry, Chemical Engineering

King Abdullah University of Science and Technology (KAUST) Jeddah, Saudi Arabia M. S. in Cheminformatics Sept. 2021 - Dec. 2022

Core Courses: Python, Machine Learning, Digital Design and Computer Architecture, Algorithms, Materials Chem- istry, Advanced Inorganic Chemistry, Advanced Physical Chemistry Virginia Tech Blacksburg, United States

Ph.D. in Computer Science, Advised by Dr. Peng Gao Jan. 2023 - Now Core Courses: Deep Learning, Natural Language Processing, Trustworthy Machine Learning HONORS AND AWARDS

IEEE Symposium on Security and Privacy Travel Grant (Top-tier Conference) 2023 CCI SWVA Cyber Innovation Scholarship 2023

SUSTech Outstanding Graduate 2022

KAUST Innovation Scholarship 2022, 2021

SUSTech Academic Scholarship 2021, 2020, 2019

National Program of Undergraduate Innovation and Entrepreneurship 2021 Second Prize of Guangdong University Student Art Gallery 2021 Advanced Individual Award of SUSTech Volunteer Association 2020 EXPERIENCE

Contrastive Learning of Molecular Structure with GNN for Drug Activity Prediction Advisor: Prof. Peng Gao, Virginia Tech Jul. 2023 – Now

• Augment molecule graphs by masking atoms and deleting bonds, facilitating the pre-training of GNNs to capture representative features.

• Employ contrastive learning to amplify the similarity of augmented versions of same molecules while diminishing the similarity among different molecules.

• Predict drug activities in datasets: BACE, BBBP, ClinTox, HIV, SIDER, Tox21, ToxCast, and Tox21-10k. Backdoor NLP Attacks Based on Text Style Transfer and Sentence Structures Advisor: Prof. Peng Gao, Virginia Tech Jan. 2023 – Now

• Explore trustworthy NLP, which includes adversarial attacks, backdoor attacks and different defenses in NLP domain.

• Use linguistic style motivated dynamic backdoor attack to transfer style of text into lyrics, and test attack success rate and accuracy with different models: Text CNN, Bert, LSTM.

• Apply sentence structures backdoor attack which uses sentence structures as triggers. Generate poisoned data with poison rate 20 and run with Bert-transfer and LSTM to evaluate loss, attack success rate and accuracy. Machine Learning in Predicting Medical Reaction Yield in C–N Cross-coupling for Drug Discovery Advisor: Prof. Limin Huang, SUSTech; Prof. Xin Gao, KAUST Feb. 2022 – Nov. 2022

• Collect reaction data from the high-throughput experimental dataset with four variables: reactants, Pd catalysts, additives, and bases. Use SMILES string to represent molecules and encode with RDK molecular fingerprint.

• Build seven machine learning models and evaluate their prediction accuracy and RMSE value. The highest accuracy is 0.94 for Random Forest classifier.

• Reach over 70 percent of prediction accuracy with improved performance, which indicates the descriptors can effectively capture and represent molecules.

Computational Chemistry for Calculating Reactivity of Pd-indenyl Complexes Bearing Phosphine Ligands

Advisor: Prof. Luigi Cavello, KAUST Sep. 2021 – Feb. 2022

• Verify and explain the experimental data from quantum chemistry and chemical model simulation with reaction rate constant. Change ligands to analyze effect of configurations with different energy and stability.

• Calculate energy of reactant, transition state, and product. Electron density and steric hindrance will directly affect the energy of the reaction intermediate, which will influence the activation energy and the reaction rate.

• Methodology: Gaussian09, M06/TZVP//PBE0-D3/SVP, Solvent Chloroform. PROFESSIONAL SERVICES

External reviewer for ACM CCS 2023.

TEACHING EXPERIENCE

Teaching assistant in Virginia Tech, CS-4804 Introduction to Artificial Intelligence Volunteer teacher in Lianping in 2019 summer.

LEADERSHIP

Project leader of national undergraduate innovation and entrepreneurship program. TECHNIQUE SKILLS

Programming: Proficient in Python, Pytorch, Java, RDKit Analysis and presentation software: Excel, PowerPoint, Word, Origin, Keynote Language: English, Mandarin

EQUIPMENT SKILLS

H-NMR, C-NMR, FTIR, Mass Spectrometry, Infrared spectrometer, XRD, UV-vis Spectrometer, Fluorescence Spec- trometer, Abbe Refractometer, HPLC, Gel Permeation Chromatography, LC-20AT, Graphite Furnace Atomic Ab- sorption Spectrometer, SEM, TEM, HPLC, DFT

SELF EVALUATION

Problem Solving, Innovative Thinking, Strong Execution, Responsible, Detail-driven, Collaborative, Effective Com- munication, Leadership

Contact this candidate