Batch-sensitized preprocessing of data to preserve genuine biological signs in analysis of multi-omics medical data: Research

In the study “Confounding interactions between missing value imputation and batch effect correction in multi-omics data integration” lead by Asst. Prof Wilson Goh, researchers are proposing that by introducing batch-aware data processing methods, it would ensure that biomedical discoveries are based on real biological signals rather than technical noise, strengthening the path toward reliable precision medicine.

 Modern medical research increasingly uses multi-omics data — large biological datasets that include DNA, proteins, and other molecules — to better understand diseases. Integrating these data complicates the analysis due to missing values such as having some biological measurements not detected and also batch effects when imputation is performed before batch correction. Batch-specific data may inadvertently be baked into the data or converting them into irreversible intra-sample noise that later correction algorithms cannot recognize or remove ^1.

Imputation of batch-wide missingness, or Batch Effect Associated Missingness (BEAMs), creates false class artifacts and spurious statistical significance ².

 The impact of this study in healthcare and medicine research is potentially more reliable biomarker discovery and safer clinical translation.

 Possible research directions include: Improved data processing algorithms, scalable multi-omics integration and AI-driven precision medicine

C-AIM Lead:

GOH Wen Bin Wilson Chief Data Scientist, Centre of AI in Medicine Asst. Professor of Biomedical Informatics Lee Kong Chian School of Medicine Nanyang Technological University

 

Related Research Domain: Others, Outcome studies

Keywords: Multi-Omics Integration, Missing value imputation, Batch effect correction

 

Reference:

1. Hui, H. W. H., Chan, W. X., & Goh, W. W. B. (2025). Assessing the impact of batch effect associated missing values on downstream analysis in high-throughput biomedical data. Briefings in Bioinformatics, 26(2), bbaf168. https://doi.org/10.1093/bib/bbaf168
2. Hui, H. W. H., Kong, W., Peng, H., & Goh, W. W. B. (2023). The importance of batch sensitization in missing value imputation. Scientific Reports, 13(1), 3003. https://doi.org/10.1038/s41598-023-30084-2