From Citation to Application: A Usage-Based Assessment of Pathway Analysis Methods in Multi-Omics Research

Pathway analysis tools are essential for interpreting high-dimensional multi-omics data. However, while many tools have been developed, their practical usage in applied research pipelines remains unclear, since tools are often cited without being actually used. Here, we systematically evaluated the practical adoption of 28 publicly available software tools designed for multi-omics pathway analysis. We analyzed 18,447 open-access full-text articles published before 2025 and downloaded the corresponding full-texts from the PubMed database. Citations were classified based on their appearance in different sections of each article via the JATSdecoder package. Tools were considered “used” if cited in methodological sections. We report that ClusterProfiler and Pathview are the most frequently used tools, applied in over 80% of citing research articles. Other tools, such as ReactomeGSA and PARADIGM, have high citation counts but were used in fewer than 50% and 20% of cases, respectively. Overall, citation frequency did not reliably reflect practical usage. Our findings highlight a discrepancy between citation and practical adoption of multi-omics pathway analysis tools, emphasizing the need for usage-based evaluation metrics to inform tool selection in bioinformatics.