Work

FSBLUP: a novel strategy of fusion similarity matrix construction via optimally integrating intermediate omics data to enhance genomic prediction

Year: 2026
Type: article
Abstract: The increasing availability of multi-omics data is promising in enhancing genomic prediction in breeding and human genetics. However, integrating multi-omics data into genomic prediction models remains challenging due to complex relationships between omics layers and phenotypic outcomes. We propose Fusion Similarity Best Linear Unbiased Prediction (FSBLUP), a novel strategy that integrates genomic and intermediate omics data using a unified similarity matrix approach. FSBLUP systematically estimates how different omics layers contribute to phenotypic variation via machine-learning-optimized parameters that capture underlying genetic architecture of complex traits. FSBLUP demonstrates greater predictive accuracy than existing methods, as validated through theoretical and practical evaluations.
Language: English
Subfield: Genetics
Open Access status: gold
Estimated APC paid: $5,490
OpenAlex ID: W7124136809
FWCI: 8.683
Cites: 63
Cited by: 5
Related to: 10