Choosing the right software for RNA sequencing (RNA-seq) analysis is as important as designing the experiment correctly. Recent publications [1, 2] demonstrated this, using different biological models to study the effects of minimal radiation exposure. The studies revealed that the use of different software packages for genome mapping, count normalization, and statistical calculation of differentially expressed genes (DEGs) led to inconsistent results. Both the number of DEGs and changes in gene expression (fold change; FC) varied considerably. Using default software parameters, CLC Genomics yielded the highest number of DEGs and FCs in 12 out of 14 comparisons, while DNAstar-D (DESeq2 in the DNAstar pipeline) was consistently the most conservative. Striking differences of up to 3 and 4 orders of magnitude in the number of DEGs and FCs, respectively, were observed between the two pipelines.
As reliable and consistent results are essential for drawing scientific conclusions, these studies highlighted the importance of selecting the appropriate software for data analysis. Whether to use commercial off-the-shelf software or to invest resources in building an in-house bioinformatics solution is another significant decision. The authors noted that many scientists, like themselves, who are not experts in bioinformatics, prioritize their resources on experimental design and biological knowledge rather than on pipeline development and maintenance. A software solution that meets both scientific requirements and produces reliable results that are comparable from lab to lab is in high demand.
PanHunter, Evotec’s proprietary multi-omics analysis platform, uses STAR aligner to map transcripts to the reference genome and DESeq2 (median of ratios normalization) for DEG analysis.
These software packages, which maintain statistical power and stability across sample sizes, produce results that are modest compared to those tested ones [2]. By default, PanHunter uses 0.01 as the false discovery rate (FDR) cutoff for DEG calculation. However, both FDR and FC cutoffs can be easily adjusted in the graphical user interface, allowing scientists to study the effects of these criteria and optimize the parameters for the dataset being analyzed. Analysis results can be interactively explored and easily retrieved, revised, and shared with internal and external collaborators. Moreover, PanHunter provides a well-maintained, scalable, and secure environment to accelerate omics analysis at all levels. Combined with the power to perform reproducible and trustworthy analyses, the flexibility to optimize statistical parameters, and the ability to share analysis results, PanHunter enables users to generate consensus results while discovering new insights, novel biomarkers, and drug targets.
In summary, RNA-Seq is a powerful tool for studying gene expression and identifying novel drug targets. To unlock the underlying biological insights, the selection of appropriate data analysis software is essential. PanHunter, Evotec’s proprietary multi-omics analysis platform, uses peer-reviewed algorithms, STAR aligner and DESeq2, combined with well-established analysis parameters to ensure reproducible and trustworthy results. In addition, the user-friendly interface allows users to interactively explore data, optimize statistical parameters, and share analysis results on the fly. PanHunter enables researchers to prioritize their scientific work and thus advances cutting-edge interdisciplinary research.
Laren more about PanHunter: www.panhunter.com
Author: Yi-Tse Liu-Grabitz
References
-
Thawng CN, Smith GB. A transcriptome software comparison for the analyses of treatments expected to give subtle gene expression responses. BMC Genomics 23, 452 (2022) [article].
-
Thawng, C.N., Smith, G.B. Transcriptome software results show significant variation among different commercial pipelines. BMC Genomics 24, 662 (2023). [article]
