KnowSeq R/bioc package: Beyond the traditional RNA-seq pipeline. A breast cancer case study.

Abstract

The number of gene expression analyses has grown exponentially over the last years. The main triggers of this increase are the reduction in the sequencing cost per sample and the technological advances, specially in the computing scope. Those analyses generally involve a number of steps. Firstly, a raw samples alignment and a quality analysis are needed. After that, a Differentially Expressed Genes (DEGs) extraction and a subsequent gene enrichment can be performed. The development of intelligent predictive tools results essential in bioinformatics given that there exists a real need of assistance for decision-making systems towards precision medicine. Therefore, KnowSeq incorporates novel steps of feature selection and classifier design in the traditional RNA-seq pipeline. No tool exists in the research community that achieves this complete RNA-seq analysis, encapsulating all those steps in one single tool. In order to show the functionalities provided by the general pipeline designed for the KnowSeq package, an application to a real problem is presented. Concretely, an analysis of a breast cancer set of patients collected from the controlled repository GDC portal is performed, keeping paired samples between tumour and control. As results show, KnowSeq achieves extracting more relevant biological knowledge related with breast cancer from the RNA raw data acquisition. KnowSeq is available through Bioconductor. KnowSeq R/bioc package is born with the purpose of providing an integrative tool, containing the necessary steps to address complex RNA-seq analyses in a modular and flexible way. In this paper a breast study case is addressed with KnowSeq, obtaining outstanding results and demonstrating the validity of KnowSeq to carry out gene expression analyses.

Towards Improving Skin Cancer Diagnosis by Integrating Microarray and RNA-Seq Datasets

“…Each of the experimental steps of this proposed pipeline will be subsequently addressed in the following subsections. For reproducibility purposes, this methodological approach can be run step-by-step by using different functionalities published under the KnowSeq R/Bioc package [38] together with several R scripts which have been included in the next repository: https://github.com/jmggugr/skca-transcriptomics-integration/…”