Brain-wide interpolation and conditioning of gene expression in the human brain using Implicit Neural Representations

2025-06-11

AI Summary

To address the low spatial resolution of whole-brain gene expression maps arising from sparse sampling in human spatial transcriptomics, this paper introduces implicit neural representations (INRs) for the first time to reconstruct whole-brain gene expression patterns. We propose a differentiable, continuous, and conditionally controllable voxel-level expression modeling framework. Using the Allen Human Brain Atlas and the Abagen benchmark, we jointly model 100 Alzheimer’s disease (AD)-risk genes across brain regions and genes. Compared with conventional interpolation methods, our approach significantly improves fidelity of expression distribution and recovery of fine-grained spatial structure, generating quantitative whole-brain expression maps at 1 mm³ resolution. Key contributions include: (i) the first systematic application of INRs to whole-brain spatial transcriptomics; (ii) support for gene-specific conditional generation and cross-gene generalization; and (iii) provision of an interpretable, computationally tractable expression prior model to facilitate mechanistic investigation of AD.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Representation Learning
• Humans and AI: Brain-Sensing and Analysis

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

In this paper, we study the efficacy and utility of recent advances in non-local, non-linear image interpolation and extrapolation algorithms, specifically, ideas based on Implicit Neural Representations (INR), as a tool for analysis of spatial transcriptomics data. We seek to utilize the microarray gene expression data sparsely sampled in the healthy human brain, and produce fully resolved spatial maps of any given gene across the whole brain at a voxel-level resolution. To do so, we first obtained the 100 top AD risk genes, whose baseline spatial transcriptional profiles were obtained from the Allen Human Brain Atlas (AHBA). We adapted Implicit Neural Representation models so that the pipeline can produce robust voxel-resolution quantitative maps of all genes. We present a variety of experiments using interpolations obtained from Abagen as a baseline/reference.

Problem

• Interpolate gene expression data in human brain
• Create voxel-level spatial maps of genes
• Analyze spatial transcriptomics using neural representations

Innovation

• Implicit Neural Representations for gene interpolation
• Voxel-resolution mapping from sparse microarray data
• Non-linear algorithms for spatial transcriptomics analysis