TemSOMap

Mapping lineage-resolved scRNA-seq data with spatial transcriptomics

TemSOMap: Mapping lineage-resolved scRNA-seq data with spatial transcriptomics

Status: Successfully concluded, submitted and under review

Role: Research Assistant, Co-author

Institution: Prof. Xiuwei Zhang Lab, College of Computing, Georgia Institute of Technology

Duration: April 2024 – September 2025 (Remote collaboration)

Project Overview

TemSOMap addresses a fundamental challenge in spatial transcriptomics analysis: how to accurately map single-cell RNA sequencing data with lineage information to spatial transcriptomics data. This project develops a sophisticated machine learning framework that considers both gene expression patterns and lineage relationships to create more accurate cell-to-spot mapping.

Technical Approach

The methodology employs advanced machine learning techniques:

Core Algorithm:

  • Loss Function Optimization: Development of a comprehensive loss function that balances expression similarity and lineage constraints
  • Cell-to-Spot Mapping: Inference of probabilistic mapping matrices between single cells and spatial spots
  • Lineage Integration: Novel incorporation of lineage tracing information into spatial mapping

Machine Learning Framework:

  • Expression-based similarity metrics
  • Lineage-aware constraint incorporation
  • Optimization algorithms for large-scale mapping problems

My Contributions

As a research assistant and co-author, my responsibilities included:

Downstream Analysis:

  • Development and implementation of comprehensive downstream analysis pipelines
  • Validation of mapping results using biological knowledge and independent datasets
  • Creation of visualization tools for mapping quality assessment

Benchmarking:

  • Partial development of benchmark datasets for method comparison
  • Performance evaluation against existing mapping methods
  • Statistical analysis of mapping accuracy and reliability

Manuscript Preparation:

  • Partial contribution to manuscript writing and revision
  • Figure preparation and data presentation
  • Literature review and method comparison

Collaborative Research Experience

This project provided valuable experience in:

  • Remote International Collaboration: Effective communication and coordination with US-based research team
  • Method Development: Participation in algorithm design and implementation processes
  • Scientific Rigor: Learning best practices in computational biology research methodology

Current Status

The project has been successfully completed and the manuscript has been submitted to a peer-reviewed journal. The work is currently under review, with revisions expected based on reviewer feedback.

Expected Impact

TemSOMap will contribute to the field by:

  • Improved Spatial Analysis: More accurate integration of single-cell and spatial transcriptomics data
  • Lineage-Aware Methods: Novel consideration of developmental relationships in spatial mapping
  • Benchmarking Standards: Establishment of evaluation criteria for mapping method assessment

Submission: Pan X, Danies-Lopez A, Chen Y, Zhang X*. Mapping lineage-resolved scRNA-seq data with spatial transcriptomics using TemSOMap. *Submitted to PSB2026 conference*

References