Abstract
Rank aggregation (RA) is the process of consolidating disparate rankings into a single unified ranking. It holds immense potential in the field of genomics. RA has applications in diverse research areas, such as gene expression analysis, meta-analysis, gene prioritization, and biomarker discovery. However, there are many challenges in the application of the RA approach to biological data, such as dealing with heterogeneous data sources, rankings of mixed quality, and evaluating the consolidated rankings. In this review, we present an overview of the existing RA methods with an emphasis on those that have been tailored to the complexities of genomics research. These encompass a broad range of approaches, from distributional and heuristic methods to Bayesian and stochastic optimization algorithms. By examining these techniques, we aim to equip researchers with the background knowledge needed to navigate the intricacies of RA in genomics data integration effectively. We review the practical applications to highlight the relevance and impact of RA methods in advancing genomics research. As the field continues to evolve, we identify open problems and suggest future directions to enhance the effectiveness of rank aggregation in genomics, by addressing the challenges related to data heterogeneity, single-cell omics and spatial transcriptomics data, and the development of clear and consistent evaluation methods. In summary, RA stands as a powerful tool in genomics research, which can offer deeper insights and more comprehensive data integration solutions.