Incorporating Omics Data in Genomic Prediction.

Area of Science:

• Genomics
• Bioinformatics
• Animal Breeding

Background:

• Genomic prediction models utilize DNA marker data to predict breeding values.
• Integrating diverse omics data (e.g., transcriptomics, proteomics) offers potential for improved prediction accuracy.
• Existing genomic prediction methodologies provide a foundation for omics data integration.

Purpose of the Study:

• To explore the rationale behind incorporating multi-omics data into genomic prediction.
• To review existing literature on the performance of omics-enhanced prediction models.
• To identify potential challenges and best practices for applying these integrated methods in breeding programs.

Main Methods:

• Literature review of omics-enhanced genomic prediction studies.
• Discussion of statistical method transferability from genomic to general omics data.
• Emphasis on the importance of variable standardization within omics relationship matrices.

Main Results:

• Omics data integration into genomic prediction is motivated by the potential for increased accuracy.
• Statistical frameworks for genomic data analysis are adaptable to broader omics datasets.
• Variable standardization is a critical consideration for omics relationship matrices, potentially more so than for SNP-based genomic relationship matrices.

Conclusions:

• The integration of multi-omics data holds promise for advancing genomic prediction in breeding.
• Statistical methodologies are adaptable, but careful data preprocessing, particularly standardization, is essential for optimal performance with omics data.
• Awareness of potential pitfalls is necessary for successful implementation in practical breeding applications.