Kent State University AI Lab

Department of Computer Science

Principal Investigator: Dr. Yuzhi Guo

We design and apply machine learning, computer vision, and deep learning techniques to solve practical problems, spanning drug discovery, medical image processing, and bioinformatics.

Research Areas

Selected publications per topic. See the PI homepage for the full list.

Large Language Models & Guideline-Driven Reasoning

Teaching models to follow explicit written rules: turning guidelines into environments a model can act in, optimizing prompts for expert annotation, and grounding multimodal reasoning in domain protocols.

  • ACL'26Guidelines as Environments: A World Model Approach to Rule Following
  • AAAI'26Learning from Guidelines: Structured Prompt Optimization for Expert Annotation Tasks
  • CVPR'26Universal Guideline-Driven Image Clustering via a Hybrid LLM Agent
  • MIDL'26Guideline-Informed MLLM Reasoning for Pathology-Aware Post-Operative Prostate CTV Segmentation

Multimodal Representation Learning

Aligning and fusing heterogeneous sources — images, text, molecules, clinical records — into shared representations, with attention to the geometry of the embedding space and to what happens when modalities are missing.

  • ICLR'26GRAM-DTI: Adaptive Multimodal Representation Learning for Drug-Target Interaction Prediction
  • CVPR'26Hyperbolic Gramian Volumes for Multimodal Alignment
  • NeurIPS'25TRIDENT: Tri-Modal Molecular Representation Learning with Taxonomic Annotations and Local Correspondence
  • ECCV'26Histopathology Multi-modal Embedding for Pathology Composed Retrieval
  • ICCV'25Zero-Shot Composed Image Retrieval via Dual-Stream Instruction-Aware Distillation
  • MIDL'26Heterogeneous Aligned Fusion for Survival Prediction with Missing Modalities

Medical Imaging & Computational Pathology

Whole-slide image analysis, nuclei segmentation, survival prediction and screening from medical images, with a focus on learning under weak labels and on predictions clinicians can interpret.

  • IEEE TNNLS'25Segment Any Cell: A SAM-based Auto-prompting Fine-tuning Framework for Nuclei Segmentation
  • NPJ Dig. Med.'25Multi-modal Analysis of Whole Slide Images in Colorectal Cancer
  • MICCAI'24PathM3: A Multimodal Multi-Task Multiple Instance Learning Framework for Whole Slide Image Classification and Captioning
  • MICCAI'25HAGE: Hierarchical Alignment Gene-Enhanced Pathology Representation Learning with Spatial Transcriptomics
  • MICCAI'25Text-Guided Multi-Instance Learning for Scoliosis Screening via Gait Video Analysis
  • MIDL'26Guideline-Informed MLLM Reasoning for Pathology-Aware Post-Operative Prostate CTV Segmentation
  • ISBI'25Contrastive Pretraining for Computational Pathology With Visual Language Models
  • Front. Med.'25Abnormality-aware Multimodal Learning for WSI Classification
  • ACM BCB'24MFMF: Multiple Foundation Model Fusion Networks for Whole Slide Image Classification
  • ICCV'21Exploring Robustness of Unsupervised Domain Adaptation in Semantic Segmentation

Gene Function & Genomics

Language models and graph methods for genomic sequence and gene annotation: motif-oriented DNA pretraining, Gene Ontology-informed representations, and predicting the phenotypic consequences of gene knockouts.

  • AAAI'26GenePheno: Interpretable Gene Knockout-Induced Phenotype Abnormality Prediction from Gene Sequences
  • AAAI'25GoBERT: Gene Ontology Graph Informed BERT for Universal Gene Function Prediction
  • BioMedInform.'24Advancing DNA Language Models through Motif-Oriented Pre-training with MoDNA
  • ACM BCB'22MoDNA: Motif-Oriented Pre-training for DNA Language Model
  • MICCAI'25HAGE: Hierarchical Alignment Gene-Enhanced Pathology Representation Learning with Spatial Transcriptomics

Protein Modeling

Representation learning for protein sequence and structure — secondary structure and structural property prediction, self-supervised pretraining from tertiary structures, and enhancing evolutionary profiles — together with the immune receptor side: BCR and TCR repertoires, antigen binding affinity and B-cell epitope prediction.

  • Nat. Cancer'25Profiling Antigen Binding Affinity of B Cell Repertoires in Tumors by Deep Learning
  • Brief. Bioinform.'24GTE: A Graph Learning Framework for Prediction of T-Cell Receptors and Epitopes Binding Specificity
  • ACM BCB'24AlphaEpi: Enhancing B Cell Epitope Prediction with AlphaFold 3
  • AAAI'22Self-supervised Pretraining for Protein Embeddings Using Tertiary Structures
  • IEEE BIBM'20Protein Ensemble Learning with Atrous Spatial Pyramid Networks for Secondary Structure Prediction
  • RECOMB'20Bagging MSA Learning: Enhancing Low-Quality PSSM with Deep Learning for Accurate Protein Structure Property Prediction
  • IEEE BIBM'20WeightAln: Weighted Homologous Alignment for Protein Structure Property Prediction
  • J. Comput. Biol.'21EPTool: A New Enhancing PSSM Tool for Protein Secondary Structure Prediction

Molecular Representation and Property Prediction

Pretraining and robust training for molecular property prediction, including learning from limited labels and handling the distribution shift that comes with new chemical space.

  • ACM BCB'19SMILES-BERT: Large Scale Unsupervised Pre-training for Molecular Property Prediction
  • ICLR'26GRAM-DTI: Adaptive Multimodal Representation Learning for Drug-Target Interaction Prediction
  • NeurIPS'25TRIDENT: Tri-Modal Molecular Representation Learning with Taxonomic Annotations and Local Correspondence
  • J. Comput. Biol.'24Towards Robust Self-training for Molecular Biology Prediction Tasks
  • IEEE BIBM'21Gradient-Norm Based Attentive Loss for Molecular Property Prediction
  • IEEE BIBM'20Improving Molecular Property Prediction on Limited Data with Deep Multi-Label Learning
  • ICML Workshop'23CoSSL: A Context-based Semi-Supervised Framework for Molecular Property Prediction

Work in Progress

Ongoing projects, with internal notes linked where available.

Gene-perturbation

In progress

Integrating multimodal gene representations for predicting transcriptional responses to single and combinatorial gene perturbations.

CRISPR perturbationRetrievalReplogleUnseen genes

KEGG

In progress

Learning transferable molecular relationships from biological pathway graphs for downstream interaction prediction across pathways and databases.

Pathway graphsNext-KO predictionLeakage-free splits

Shared-Residual World Models

In progress

A multimodal world model that splits each modality into a shared summary plus a modality-specific residual instead of forcing one aligned latent. Preliminary ALFRED results: better next-state prediction and native missing-modality inference, with no alignment loss.

World modelsMultimodal ALFREDMissing modality

Modality-Aware Fusion for AD Conversion Prediction

In progress

Predicting progression from MCI to Alzheimer's disease using ADNI imaging, PET, protein, genetic and clinical data. Our modality-aware fusion model matches the strongest tabular baseline on complete data and pulls significantly ahead once input modalities go missing — the setting it would actually be deployed under.

ADNIMultimodal Alzheimer'sMissing modality
Links marked ↗ OneDrive require access. Only Kent State University email accounts (@kent.edu) can request access. Download the HTML file and open it in a browser to read it.

People

Principal Investigator

Ph.D. Students

Mascot

Bunny, the lab mascot

Bunny

Lab Mascot

Opportunities

Welcome to join our team!

We are always looking for motivated students and collaborators who are interested in machine learning, computer vision, and AI for biology and medicine. If our work sounds exciting to you, we would love to hear from you.

Contact

Email
yguo30@kent.eduDr. Yuzhi Guo, Principal Investigator
Address
Department of Computer Science
Mathematical Sciences Building (MSB)
1300 University Esplanade
Kent State University, Kent, OH 44242
Students
Prospective students: see Opportunities.
Mathematical Sciences Building (MSB), Kent State University Open in Google Maps