Bayesian factor models play a central role in the analysis of high-dimensional and structured data arising in bioinformatics and biostatistics. This special session aims to highlight recent methodological advances that extend classical factor analysis to address key challenges posed by modern biomedical data, including complex dependence structures, heterogeneity, and interpretability.

Particular emphasis will be placed on structured and sparse Bayesian factor models, with attention to identifiability, invariance, and dynamic behavior. The session will feature recent developments in informed identifiability for sparse factor models, tensor and dynamic factor decompositions, rotationally invariant formulations, and meta-analysis data approaches. Contributions will also illustrate how these models support principled data integration, causal inference, and uncertainty quantification in complex settings.

Applications motivating these methodological advances include neuroimaging meta-analysis, cross-species data integration, public health intervention evaluation, and other high-dimensional biomedical problems. By bringing together complementary perspectives on structure, sparsity, and identifiability in Bayesian factor modeling, this session aims to foster discussion on robust and interpretable latent variable modeling strategies.

The expected impact is to strengthen the connection between advanced Bayesian methodology and pressing applied problems in bioinformatics and biostatistics, promoting the development of scalable, principled, and scientifically meaningful factor modeling approaches.