AUTHOR=Mallik Saurav , Seth Soumita , Si Amalendu , Bhadra  Tapas , Zhao Zhongming 

TITLE=Optimal ranking and directional signature classification using the integral strategy of multi-objective optimization-based association rule mining of multi-omics data

JOURNAL=Frontiers in Bioinformatics

VOLUME=3

YEAR=2023

URL=https://www.frontiersin.org/journals/bioinformatics/articles/10.3389/fbinf.2023.1182176

DOI=10.3389/fbinf.2023.1182176

ISSN=2673-7647

ABSTRACT=<p><bold>Introduction:</bold> Association rule mining (ARM) is a powerful tool for exploring the informative relationships among multiple items (genes) in any dataset. The main problem of ARM is that it generates many rules containing different rule-informative values, which becomes a challenge for the user to choose the effective rules. In addition, few works have been performed on the integration of multiple biological datasets and variable cutoff values in ARM.</p><p><bold>Methods:</bold> To solve all these problems, in this article, we developed a novel framework <italic>MOOVARM</italic> (multi-objective optimized variable cutoff-based association rule mining) for multi-omics profiles.</p><p><bold>Results:</bold> In this regard, we identified the positive ideal solution (<italic>PIS</italic>), which maximized the profit and minimized the loss, and negative ideal solution (<italic>NIS</italic>), which minimized the profit and maximized the loss for all gene sets (item sets), belonging to each extracted rule. Thereafter, we computed the distance (<italic>d</italic> +) from PIS and distance (<italic>d</italic> −) from NIS for each gene set or product. These two distances played an important role in determining the optimized associations among various pairs of genes in the multi-omics dataset. We then globally estimated the relative closeness to <italic>PIS</italic> for ranking the gene sets. When the relative closeness score of the rule is greater than or equal to the pre-defined threshold value, the rule can be considered a final resultant rule. Moreover, <italic>MOOVARM</italic> evaluated the relative score of the rule based on the status of all genes instead of individual genes.</p><p><bold>Conclusions:</bold><italic>MOOVARM</italic> produced the final rank of the extracted (multi-objective optimized) rules of correlated genes which had better disease classification than the state-of-the-art algorithms on gene signature identification.</p>