Interpretable representation learning
VAE-family models, geometry-aware latent spaces, and stable embeddings designed for clearer clustering, visualization, and biological interpretation.
iVAE
LiVAE
CCVGAE
Identity layer for research, software, and public tools
Zeyu Fu 付泽宇
Ph.D. Candidate · Single-Cell Computational Biology · Machine Learning
I build machine learning systems for single-cell genomics, spanning interpretable representation learning, neural ODEs, reinforcement learning, and graph-based modeling for cell-state dynamics, trajectory inference, and benchmark-ready analysis.
8
peer-reviewed papers highlighted on this homepage
8
published software projects with code or package distribution
4
public graph surfaces linked directly from the identity layer
Choose a route
See full public graph section
Choose the right public route from the homepage
Use the identity layer for orientation, then jump into the right public surface based on whether you need discovery, benchmarks, project pages, or a focused utility.
Hub
Primary Route
SCPortal
Canonical public discovery hub for datasets, benchmarks, models, and related tools.
Open SCPortal →
Utility
Primary Route
mRNA Intersection
Interactive mRNA intersection analysis and visualization tool.
Launch →
Microsite
Secondary Route
LAIOR Benchmarks
Focused benchmark destination for model summaries, datasets, and metric detail pages.
Open LAIOR Benchmarks →
Public Surface
Secondary Route
iAODE Pages
Public-facing iAODE destination for datasets, explorer pages, and project overview content.
Open iAODE Pages →
Flagship journey
Identity → discovery → benchmark proof
The flagship triad gives visitors a deliberate path: verify the researcher and publications here, use SCPortal to choose the right single-cell route, then open LAIOR Benchmarks for focused model, dataset, and metric inspection.
Academic provenance
Public graph continuity
Benchmark microsite
01
02
Homepage: identity root
Use this page for name disambiguation, publication context, software provenance, and trusted profile links before following a tool route.
SCPortal: discovery hub
Use the hub when the task is still exploratory: datasets, models, continuity exploration, benchmark context, and related destinations.
Open hub → 03LAIOR: benchmark microsite
Use the focused microsite once the question is benchmark-specific: compare models, inspect datasets, and understand metrics.
Open benchmarks →About
Representation learning, dynamics, and decision-making methods for single-cell sequencing data.
I develop machine learning methods for single-cell sequencing data (scRNA-seq and scATAC-seq), with published work spanning variational autoencoders, neural ODEs, reinforcement learning, graph neural networks, and hyperbolic geometry applied to cell fate analysis and representation learning.
Single-Cell GenomicsTrajectory InferenceVAENeural ODEReinforcement LearningGNNHyperbolic GeometryContrastive LearningClusteringFlow Matching
What this homepage helps visitors do
Understand the research arc
A concise read of recent publications, open-source releases, and the modeling themes connecting them.
Find the right public destination
The homepage routes visitors toward SCPortal for broad discovery or directly into a curated set of public tools.
Stay at overview level unless depth is needed
The page stays intentionally lightweight and identity-first rather than turning into a portal-scale index.
Research focus
Match focus to software
Three recurring themes across the work
A compact map of the modeling directions that connect publications, software, and public app surfaces.
Dynamics and continuum modeling
Neural ODE and graph-enhanced methods for developmental trajectories, chromatin accessibility continua, and fate-transition structure in single-cell systems.
iAODE
GNODEVAE
CODE
Decision-aware single-cell analysis
Reinforcement learning and downstream evaluation utilities for branching, lineage focus, and targeted biological workflow support.
scRL
scFocus
mRNA Intersection
Publications
# Equal contribution * Corresponding author 8 peer-reviewed papers
Open-Source Software
8 published tools (7 on PyPI)
iAODE
pip install iaode
Neural ODE-VAE for scATAC-seq benchmarking and continuum modeling.
iVAE
pip install iVAE
Interpretable VAE enhancing clustering for single-cell data.
LiVAE
pip install livae
Lorentz-regularized VAE for transcriptomic and epigenomic embeddings.
GNODEVAE
pip install gnodevae
Graph ODE-VAE with GNN-enhanced clustering and dynamics.
scRL
pip install scrl-fatedecision
RL for evaluating cell fate decisions in single-cell data.
scFocus
pip install scfocus
SAC-based branching probability detection for lineage focusing.
CODE
pip install scCODE
Correlated latent space learning for single-cell RNA continuum modeling.
CCVGAE
Array · 2026
Centroid-coupled variational graph attention autoencoder for single-cell representation learning.
Public graph
Curated Interactive Surfaces
Start with SCPortal for broad discovery, then move into a small, deliberate set of public destinations surfaced from the homepage.
Identity
Discovery
Focused workflows
Primary entry points
Secondary destinations
Also in the public graph
Hidden by design
Boundaries preserved
Local-First
Local-First boundary
Kept off the homepage because it is a local-first workspace rather than a public hosted app surface.
Landing-Only
Landing-Only boundary
Maintained as a landing-only reference surface, so the homepage does not present it as an actively browsable public tool.
Contact
For collaborations, code, profiles, and public research references.