Skip to content
View yazdanimehdi's full-sized avatar

Highlights

  • Pro

Block or report yazdanimehdi

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
yazdanimehdi/README.md

Hi, I'm Mehdi Yazdani

AI Researcher • Ph.D. Candidate • Computational Biology Enthusiast


🧬 About Me

I'm a Ph.D. candidate in Computer Science at the University of Central Florida, working at the intersection of machine learning, computational drug discovery, and bioinformatics. My research focuses on building robust, symmetry-aware models for biological sequence modeling, particularly in the context of mRNA, drug-target interaction, and fairness in AI systems.


🧪 Research Highlights

  • HELM (ICLR 2025): Developed the first hierarchical mRNA language model, improving antibody prediction by 8%.
  • FragXsiteDTI (NeurIPS 2024): Transformer-based model for interpreting drug-target binding sites.
  • FairBiNN (NeurIPS 2024): Bilevel optimization for balancing accuracy and fairness in neural networks.
  • DeepDrugDomain: Toolkit for standardizing datasets and benchmarks in drug discovery tasks.
  • Equi-mRNA (under review): Group-theoretic model for codon embeddings and equivariant mRNA generation.

🛠️ Technical Skills

  • Languages: Python, C++, JavaScript, Dart, SQL, Solidity
  • Frameworks: PyTorch, TensorFlow, Hugging Face Transformers
  • Specialties: Group Theory in ML, Geometric Deep Learning, Language Modeling
  • Tooling: Kubernetes, Docker, AWS, Slurm, Triton, CUDA

📍 Current & Upcoming

  • 💼 AI Research Scientist Intern @ Microsoft Research (Bio-LLMs)

📚 Publications & Media

  • Papers at NeurIPS, ICLR, RECOMB, and Briefings in Bioinformatics
  • Featured on ABC News and UCF Today for drug repurposing work during the COVID-19 pandemic

📫 Connect With Me


🧠 Building the next generation of biological AI models, one codon at a time.

Pinned Loading

  1. DeepDrugDomain DeepDrugDomain Public

    DeepDrugDomain: A versatile Python toolkit for streamlined preprocessing and accurate prediction of drug-target interactions and binding affinities, leveraging deep learning for advancing computati…

    Python 23 8

  2. AttentionSiteDTI AttentionSiteDTI Public

    This is The repository for Paper "AttentionSiteDTI: Attention Based Model for Predicting Drug-Target Interaction Using Graph Representation of Ligands and 3D Structure of Protein Binding Sites"

    Python 44 9

  3. FragXsiteDTI FragXsiteDTI Public

    Drug Target Interaction Prediction Using Protein Binding Sites And Drug Fragments

    Python 11 2

  4. FairBiNN FairBiNN Public

    Fair Bilevel Neural Network (FairBiNN): On Balancing fairness and accuracy via Stackelberg Equilibrium

    Python 9 2