Applying the DeepSeqDock Harmonization Framework on Pseudomonas Aeruginosa Bacterial Biofilm Data

In the modern era, data availability has exponentially increased. From different universities, institutions, and laboratories across the globe, the wealth of information brings potential for integration and downstream analysis. This paper focuses specifically on the integration of RNA-sequencing data using data harmonization, a technique that reduces the inter-institutional batch effects that confound data quality with non-biological factors. One such framework is the DeepSeqDock framework, which creates data harmonization pipelines through training and tuning in specific domains and is evaluated on human RNA data from the SEQC. In this study, we present an alternate application of the adjusted DeepSeqDock framework for bacterial biofilm data. By adjusting for non-human data, we show the generalization of data harmonization and the DeepSeqDock for its applications in specific bacterial domains of interest for downstream analysis.