Skip to main content

Nasy Land

About Me

Ph.D. candidate supervised by Dr. Junzhou Huang at UT Arlington. Research focus: deep learning for computational biology and multimodal learning.

16 Publications
  • 5 Journal Articles
  • 5 Conference Papers
  • 3 Preprints
  • 1 Book Chapter
  • 2 First Author
130+ Citations
  • Segment Any Cell: 61 citations
  • COVID-19 Mapping: 24 citations
  • GTE: 19 citations
  • Hyperbolic MCI: 11 citations
Nature Cancer
Profiling antigen-binding affinity of B cell repertoires in tumors by deep learning predicts immune-checkpoint inhibitor treatment outcomes
B Song, K Wang, S Na, et al.
Nature Cancer, 2025 DOI
IEEE TNNLS
Segment Any Cell: A SAM-based Auto-prompting Fine-tuning Framework for Nuclei Segmentation
S Na, Y Guo, F Jiang, H Ma, J Gao, J Huang
IEEE TNNLS, 2025 · 61 citations arXiv
MICCAI Oral
Multimodal Deep Fusion in Hyperbolic Space for Mild Cognitive Impairment Study
L Zhang, S Na, T Liu, D Zhu, J Huang
MICCAI 2023 · 11 citations Springer
IJCAI 2025
Multi-step Retrosynthesis with Reinforcement Learning
S Na, et al.
IJCAI 2025 · HeroX Semi-finalist

Research

🧬

Computational Biology

Deep learning for protein binding prediction & molecular design

Key Projects

  • TCR-pMHC Binding
    +7.1% ROC AUC Nature Cancer Book Chapter
  • BCR-Antigen (Cmai)
    0.907 AUROC, clinical ICI response prediction
  • Retrosynthesis
    +7.1% Top-1 (USPTO), HeroX Semi-finalist IJCAI 2025
🔬

Foundation Models

Adapting foundation models for biomedical imaging

Key Projects

  • Segment Any Cell
    Dice 84.03/93.04 SOTA IEEE TNNLS
  • Pathology VLMs
    Multimodal WSI classification ISBI
  • Live Demo
    segment-any-cell.com
🌐

Multimodal Learning

Hyperbolic geometry for multimodal alignment

Key Projects

Scroll for more

Projects

Segment Any Cell

Interactive demo for SAM-based cell segmentation

Description
Auto-prompting framework that adapts SAM for nuclei segmentation. Achieves SOTA on multiple benchmarks with zero manual prompts.
Tech Stack
PyTorch SAM Gradio
Impact
IEEE TNNLS publication, 61+ citations

jaxrie

JAX library for hyperbolic neural networks

Description
Comprehensive library implementing hyperbolic geometry operations and neural network layers in JAX. Supports Poincaré ball and Lorentz models.
Features
Hyperbolic embeddings, Möbius operations, hyperbolic attention, geodesic computations
Tech Stack
JAX Equinox Python

nadl

Deep learning framework built on JAX and Equinox

Description
Personal deep learning framework featuring modular architecture, custom training loops, and utilities for research experiments.
Features
Data loaders, training utilities, model components, experiment tracking
Tech Stack
JAX Equinox Optax

Nasy's Emacs

Modern Emacs configuration for developers

Description
Comprehensive Emacs configuration with modern UX, featuring LSP integration, Git tools, and language support for Python, JAX, Haskell, and more.
Features
Tree-sitter, Native compilation, Org-mode, LSP, Magit, Vertico/Consult
Tech Stack
Emacs Lisp Org-mode

naipyext

IPython extensions for better debugging

Description
Enhanced IPython extensions with improved traceback formatting, syntax highlighting, and debugging utilities for data science workflows.
Features
Rich tracebacks, auto-reload, magic commands, Jupyter integration
Tech Stack
Python IPython

Research Archive

Collection of research code and experiments

Description
Archive of research implementations including HyperGRAM, retrosynthesis models, and various deep learning experiments.
Contents
Video-text retrieval, molecular design, multimodal learning experiments
Tech Stack
PyTorch JAX Python

Skills

Expert Python, JAX with Equinox, PyTorch, NumPy
Proficient TensorFlow with Keras, Lisp (Emacs Lisp), LLM
Familiar Haskell, JavaScript, C/C++

Publications