A comparison of RNA-Seq data preprocessing pipelines for transcriptomic predictions across independent studies

Sep 21, 2026

Uploaded by Aurora

RNA sequencing combined with machine learning enables molecular classification of cancer by building tissue-of-origin predictors from gene expression measurements. However, predictors often suffer from suboptimal performance when integrating datasets generated in different labs, due to variable data quality, inconsistent preprocessing, and unwanted noise. This work evaluates how preprocessing choices—normalization, batch effect correction, and data scaling—impact cross-study tissue-of-origin predictions using large-scale RNA-Seq datasets.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages22

Reading Time55 min

AccessFree to download

Background

Cancer tissue-of-origin prediction with RNA-Seq and machine learning

Results

Impact of normalization, batch effect correction, and data scaling

Cross-study evaluation using training and independent test sets

Conclusion

When preprocessing improves or worsens classifier performance

Van etal. BMC Bioinformatics (2024) 25:181 BMC Bioinformatics [https://doi.org/10.1](https://doi.org/10.1) 186/s12859‑024‑05801‑x RESEARCH Open Access A comparison of RNA‑Seq data preprocessing pipelines for transcriptomic predictions across independent studies Richard Van1,3, Daniel Alvarez2,3, Travis Mize4, Sravani Gannavarapu2,3, Lohitha Chintham Reddy2,3, Fatma Nasoz2,3 and Mira V. Han 1,3* *Correspondence: [mira.han@unlv.edu](mira.han@unlv.edu) 1 School of Life Sciences, University of Nevada Las Vegas, Las Vegas, NV, USA 2 Department of Computer Science, University of Nevada Las Vegas, Las Vegas, NV, USA 3 Nevada Institute of Personalized Medicine, Las Vegas, NV, USA 4 Icahn School of Medicine at Mount Sinai, Institute for Genomic Health, New York City, NY, USA Abstract Background: RNA sequencing combined with machine learning techniques has provided a modern approach to the molecular classification of cancer. Class predictors, reflecting the disease class, can be constructed for known tissue types using the gene expression measurements extracted from cancer patients. One challenge of cur rent cancer predictors is that they often have suboptimal performance estimates when integrating molecular datasets generated from different labs. Often, the quality of the data is variable, procured differently, and contains unwanted noise hampering the ability of a predictive model to extract useful information. Data preprocessing methods can be applied in attempts to reduce these systematic variations and harmonize the datasets before they are used to build a machine learning model for resolving tissue of origins. Results: We aimed to investigate the impact of data preprocessing steps—focusing on normalization, batch effect correction, and data scaling—through trial and comparison. Our goal was to improve the cross-study predictions oftissue of origin for common cancers on large-scale RNA-Seq datasets derived from thousands of patients and over a dozen tumor types. The results showed that the choice of data preprocessing operations affected the performance of the associated classifier models constructed for tissue of origin predictions in cancer. Conclusion: By using TCGA as a training set and applying data preprocessing methods, we demonstrated that batch effect correction improved performance measured by weighted F1-score in resolving tissue of origin against an independent GTEx test dataset. On the other hand, the use of data preprocessing operations worsened classification performance when the independent test dataset was aggregated from separate studies in ICGC and GEO. Therefore, based on our findings with these publicly available large-scale RNA-Seq datasets, the application of data preprocessing techniques to a machine learning pipeline is not always appropriate. Keywords: RNA-Seq, Classification, Cancer, Batch effect correction, Normalization, Data scaling © The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit [http://](http://)[ ](http://)[creativecommons.org/licenses/by/4.0/](creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver ()[. The Creative Commons Public Domain Dedication waiver (](creat

FAQs

What challenge affects cancer tissue-of-origin predictors across studies?

Datasets generated by different labs vary in quality and preprocessing, introducing systematic differences and unwanted noise that reduce the reliability of performance estimates.

Which preprocessing steps were evaluated in the study?

The study focused on normalization, batch effect correction, and data scaling as key preprocessing operations before building machine learning classifiers.

How did batch effect correction perform across independent datasets?

Batch effect correction improved tissue-of-origin prediction performance when the model was trained on TCGA and tested on an independent GTEx dataset, measured by weighted F1-score.

When did preprocessing operations worsen classification performance?

Preprocessing operations worsened classification performance when the independent test dataset was aggregated from separate studies in ICGC and GEO.

Sep 21, 2026

Uploaded by Aurora