OmicsARules: a R package for integration of multi-omics datasets via association rules mining

Area of Science:

• Bioinformatics
• Genomics
• Systems Biology

Background:

• High-throughput technologies generate large multi-omics datasets.
• Concurrent gene alterations are observed within and across omics datasets.
• Existing bioinformatics tools lack methods for assessing co-occurring aberrations across multiple regulatory levels.

Purpose of the Study:

• Introduce OmicsARules, an R-package for identifying concerted gene changes.
• Develop and implement a novel rule-interestingness measure, Lamda3, for biological pattern prioritization.
• Illustrate mechanistic connections between DNA methylation and transcription using multi-omics data.

Main Methods:

• Association rule mining framework.
• Novel rule-interestingness measure, Lamda3.
• Application to DNA methylation and RNA-seq datasets (BRCA, ESCA, LUAD).

Main Results:

• OmicsARules effectively identifies concerted gene alterations.
• Lamda3 measure demonstrates superior biological significance compared to other ranking methods.
• Demonstrated mechanistic links between methylation and transcription.

Conclusions:

• OmicsARules facilitates the exploration of concurrent patterns in omics data.
• Provides a new dimension for analyzing single or multiple omics datasets.
• Supports analysis across various sequencing platforms.