Spatial Transcriptomics
Ongoing
Current project — developing computational methods for spatial transcriptomics data to resolve tumor structure and tissue-level context in cancer.
Decoding cancer one cell at a time.
I'm a Ph.D. student in Computer Science at Florida State University, working at the intersection of machine learning and biology — using computational methods to make sense of cancer at the resolution of single cells. My research spans graph neural networks and unsupervised representation learning, copy number alteration (CNA) analysis, and spatial transcriptomics, with the broader goal of turning multi-modal biological signals into tools that help identify tumor cells, trace how cancer evolves, and eventually support clinical decisions.
Research
Graduate Researcher — Florida State University
Aug 2025 – Present
My work centers on computational methods for single-cell and spatial cancer genomics — spanning graph representation learning, tumor evolution modeling, and spatial biology. Below are the projects I've led or contributed to.
DeepMalignant
Unsupervised multimodal graph attention autoencoder for label-free tumor cell identification — completed
- Integrated scRNA-seq expression and copy number alteration (CNA) signals through graph-based representation learning.
- Designed a graph neural network architecture with multi-head attention and a contrastive (InfoNCE) objective to separate malignant from non-malignant cells without labels.
- Extended to spatial transcriptomics (Visium), recovering spatial tumor regions consistent with histopathology annotations.
- Robust across highly heterogeneous datasets, including challenging tumor microenvironments such as pancreatic PDAC.
~0.97
peak F1 score
33
datasets evaluated
4
cancer types
3
sequencing platforms
10x Genomics
inDrop
Drop-seq
Outperforms CopyKAT
Outperforms scMalignantFinder
scLongTree
Contributor
Assisted with the revision of scLongTree, a computational tool for inferring longitudinal trees from single-cell DNA sequencing data.
Publications & Preprints
Publications
Multi-modality graph representation learning for malignant cell identification from scRNA-seq using DeepMalignant
bioRxiv, 2026
P. Bhattarai, W. Yuan, H. Chi, X. M. Zhou, X. Mallory
scLongTree: an accurate computational tool to infer longitudinal trees from single-cell DNA sequencing data
bioRxiv, 2023
R. Khan, P. Bhattarai, L. Zhang, X. M. Zhou, X. Mallory
Detection of pneumonia using machine learning
ACM ICIMMI, 2023
P. Bhattarai, V. K. Kumar, B. T., R. Od
Conference Presentations
Presentations
DeepMalignant: multi-modality graph representation learning for malignant cell identification from scRNA-seq
2026
Poster presentation — Intelligent Systems for Molecular Biology (ISMB 2026)
Detection of pneumonia using machine learning
Dec 2023
5th International Conference on Information Management & Machine Intelligence (ICIMMI 2023)
Patents & Professional Service
Patents and Service
Patents
AI-based plant extract analyzing device to identify disease
Issued 2023
Co-inventor · Design Patent No. 392971-001, Government of India
Professional Service
Ad-hoc Reviewer, Nature Communications
2026 – Present
Invited to peer-review manuscripts in computational biology.
Teaching
Graduate Teaching Assistant — Florida State University
Aug 2025 – Summer 2026
Courses supported
- CGS2060 — Computer Fluency (Spring 2026, Fall 2025, Summer 2026)
- CGS2100 — Microcomputer Applications for Business/Economics (Spring 2026, Fall 2025, Summer 2026)
- Held regular office hours supporting computational problem solving, debugging, and algorithmic reasoning.
- Assessed student work against structured rubrics and delivered constructive technical feedback.
- Helped students work through projects and kept course logistics running smoothly.
Education
Education
Florida State University
Aug 2025 – Present
Ph.D. in Computer Science · Tallahassee, FL
GPA: 4.0 / 4.0 · Focus: computational biology & bioinformatics
SRM Institute of Science and Technology
2020 – 2024
B.Tech in Computer Science and Engineering · Chennai, India
CGPA: 9.53 / 10 — First Class with Distinction
Relevant coursework
COT5507 Analytic Methods for CS
CAP5540 Bioinformatics — Sequence Analysis
COT5405 Advanced Algorithms
CAP5638 Pattern Recognition
18CSC305J Artificial Intelligence
18GEO104T Computational Genomics
18CSC201J Data Structures & Algorithms
18ECO108J Embedded System Design
18NTO403T Scientific Research Principles
Skills
Skills
Machine learning & deep learning
Graph neural networks
Autoencoders
Contrastive learning
PyTorch
XGBoost
Random forests
Computational biology / genomics
scRNA-seq & scDNA-seq
Variant calling (SNV/Indel)
CNA analysis
Spatial transcriptomics
Bioinformatics tools
BWA
Samtools
GATK / Mutect2
Scanpy
Seurat
AnnData
Ensembl BioMart
Programming & infrastructure
Python (advanced)
C/C++
Bash
Linux/Unix
SLURM & HPC clusters
Git
Honors & Awards
Awards
Winning Team — SRM Smart Campus Hackathon
2023
Compex Scholarship — Government of India
2020
Contact
Get in touch
Open to conversations on single-cell genomics, graph representation learning, and computational oncology — reach out by email.
linkedin.com/in/pankajbhattarai
Tallahassee, FL