Kaito Shiku(Graduate School of Information Science and Electrical Engineering)’s paper has been accepted for Association for the Advancement of Artificial Intelligence (AAAI) 2026.
Congratulations!
Authors
Kaito Shiku, Kazuya Nishimura, Shinnosuke Matsuo, Yasuhiro Kojima, Ryoma Bise
Affiliation
Graduate School of Information Science and Electrical Engineering,
Department of Information Science and Technology
Manuscript Title
Auxiliary Gene Learning: Spatial Gene Expression Estimation by Auxiliary Gene Selection
Abstract
Spatial transcriptomics (ST) is a novel technology that enables the observation of gene expression at the resolution of individual spots within pathological tissues. ST quantifies the expression of tens of thousands of genes in a tissue section; however, heavy observational noise is often introduced during measurement. In prior studies, to ensure meaningful assessment, both training and evaluation have been restricted to only a small subset of highly variable genes, and genes outside this subset have also been excluded from the training process. However, since there are likely co-expression relationships between genes, low-expression genes may still contribute to the estimation of the evaluation target. In this paper, we propose (AGL) that utilizes the benefit of the ignored genes by reformulating their expression estimation as auxiliary tasks and training them jointly with the primary tasks. To effectively leverage auxiliary genes, we must select a subset of auxiliary genes that positively influence the prediction of the target genes. However, this is a challenging optimization problem due to the vast number of possible combinations. To overcome this challenge, we propose Prior-Knowledge-Based Differentiable Top- Gene Selection via Bi-level Optimization (DkGSB), a method that ranks genes by leveraging prior knowledge and relaxes the combinatorial selection problem into a differentiable top- selection problem. The experiments confirm the effectiveness of incorporating auxiliary genes and show that the proposed method outperforms conventional auxiliary task learning approaches.
Journal name
Association for the Advancement of Artificial Intelligence (AAAI) 2026
Relevant SDGs
SDGs 3 (Good health and well-being)
SDGs 9 (Industry, Innovation, Technology and Infrastructure)
Comments
Gene expression data play an important role in uncovering the mechanisms of diseases, but observing gene expression levels from biological tissues involves high experimental costs. To address this challenge, a research project that aims to “automatically estimate gene expression levels from images of biological tissues using image recognition models” has been accepted to a top international conference in the field of AI!
