“Quality-Control Sequencing (QCseq)”: An R package providing sample-level QC scores to assess data quality.

QCseq is an R package developed to compute and visualize quality control (QC) metrics for single-cell RNA sequencing (scRNAseq) gene expression data, providing sample-level QC scores to assess data quality.

For background, scRNAseq gene expression data lets you quantify the level of expression of different genes for individual cells. This is very useful for researchers as the pattern of gene expression can be used as a “fingerprint” of the state of a sample during the time of measurement. Details about which genes are relevant to the fingerprint (this is referred to as genes that are “differentially expressed”) can give researchers insight into a particular condition of interest or even biomarkers (Andrews & Hemberg, 2018).

As of right now, researchers must manually examine multiple independent QC metrics such as the number of detected genes, total transcriptions, and mitochondrial RNA percentage to assess sample quality, a time-consuming and subjective task. QCseq addresses this by integrating multiple QC metrics into a single, robust QC score ranging from 0 to 100. This unified score provides a more objective and reproducible measure of data quality, allowing researchers to quickly identify low-quality samples prior to any downstream analyses (e.g. differential gene expression analysis). Unlike some existing packages which only output metrics independently such as RNAseqQC (Ziebell, 2024), Seurat (Hao et al., 2023), or scater (McCarthy et al., 2017), QCseq systematically provides a unified score to reflect overall sample integrity, a novel advancement. The QCseq package was developed using R version 4.5.1 (2025-06-13), Platform: aarch64-apple-darwin20 (64-bit) and Running under: macOS Sequoia 15.6.1.

To install the latest version of the package:

install.packages("devtools")
library("devtools")
devtools::install_github("anzarlv/QCseq", build_vignettes = TRUE)
library("QCseq")

To run the Shiny app:

runQCseq()

ls("package:QCseq")
data(package = "QCseq") 
browseVignettes("QCseq")

QCseq contains 5 functions.

1. QCcompute for calculating independent quality control metrics (number of genes, number of transcripts, percentage of mitochondrial genes, and percentage of ribosomal genes) from single-cell RNA sequencing (scRNAseq) data.

2. QCscore for calculating a unified quality control score from independent QC metrics using median absolute deviation (MAD) standardization and rank-based scores.

3. QCsummary for a clear summary report for the gene expression data using the key quality control metrics extracted and the signature QC score that combines the metrics.

4. QChistogram for a publication-level histogram where x-axis is quality control score (from QCscore function) and y-axis is frequency.

5. QCpca for a plot of principal components analysis (PCA) applied to numeric QC features to reduce the data to 2 main dimensions (PCx and PCy), with the defaults being PC1 and PC2 which capture the greatest sources of variation among the cells.

The package also contains one scRNA sequencing dataset, called sample_scRNAseq. Refer to package vignettes for more details. An overview of the package is illustrated below.

The author of the package is Anzar Alvi.

The author wrote the QCcompute function, which calculates four independent QC metrics (number of genes, number of transcripts, percentage of mitochondrial genes, and percentage of ribosomal genes). Lu et al.’s (2023) research team helped the author determine the best QC metrics to extract. The QCcompute function makes use of the GetAssayData function from the Seurat (Hao et al., 2023) package to properly extract the expression matrix from the input Seurat object, and the colSums function from base R to calculate, the total number of genes, total number of transcripts, and gene percentages. ChatGPT (OpenAI, 2025) was used to set up the template for the helper function get_gene_percentage which is called in the QCcompute function. The function was completed by me after having the template.

The author wrote the QCscore function, which computes a unified QC score using median absolute deviation (MAD) standardization (Source) and rank-based scoring from 3 independent metrics: 1) number of genes, 2) number of transcripts, and 3) percentage of mitochondrial genes. The idea to use this standardization and rank-based scoring method came from Bacher et al.’s (2021) research team. The QCscore function makes use of base R packages including the stats (R Core Team, 2025) package to compute the median and MAD for different QC metrics (called in the robust_scale helper function). ChatGPT (OpenAI, 2025) was used to set up the template for the helper function robust_scale which is called in the QCscore function. The function was completed by me after having the template.

The author wrote the QCsummary function, which returns a clear summary report for the gene expression data using the key quality control metrics (n_genes, n_transcripts, percent_mit, and percent_ribo) and the signature QC score that combines the metrics. The QCsummary function makes use of the stats (R Core Team, 2025) package to derive insights (such as sd) from the metrics. No artificial intelligence tools were used for the development of this function.

The QChistogram function returns a publication-level histogram of QC signature scores for all inputted samples in the assay matrix. The QChistogram function makes use of the ggplot function from the ggplot2 (Wickham, 2016) package to plot the histogram. No artificial intelligence tools were used for the development of this function.

The QCpca function performs a principal component analysis (PCA) on key scRNAseq QC metrics. The PCA reduces the data to 2 main dimensions (PCx and PCy). Defaults are PC1 and PC2 which capture the greatest sources of variation among the cells. The function returns a ggplot2 (Wickham, 2016) gplot PCA scatter plot labeled by QC rank score. Furthermore, the QCpca function makes use of the prcomp function from the stats package.

Andrews, T. S., & Hemberg, M. (2018). Identifying cell populations with scRNASeq. Molecular Aspects of Medicine, 59, 114–122. https://doi.org/10.1016/j.mam.2017.07.002.

Bacher, R., Chu, L.-F., Argus, C., Bolin, J. M., Knight, P., Thomson, J., Stewart, R., & Kendziorski, C. (2022). Enhancing biological signals and detection rates in single-cell RNA-seq experiments with cDNA library equalization. Nucleic Acids Research, 50(2), e12–e12. https://doi.org/10.1093/nar/gkab1071.

Chang W, Cheng J, Allaire J, Sievert C, Schloerke B, Aden-Buie G, Xie Y, Allen J, McPherson J, Dipert A, Borges B (2025). shiny: Web Application Framework for R. R package version 1.12.0, https://shiny.posit.co/.

Hao, Y., Stuart, T. A., Kowalski, M. H., Choudhary, S., Hoffman, P., Hartman, A., Srivastava, A., Molla, G., Shaista Madad, Fernandez-Granda, C., & Rahul atija. (2023). Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nature Biotechnology, 42, 293–304. https://doi.org/10.1038/s41587-023-01767-y

Lu, J., Sheng, Y., Qian, W., Pan, M., Zhao, X., & Ge, Q. (2023). scRNA‐seq data analysis method to improve analysis performance. Iet Nanobiotechnology, 17(3), 246–256. https://doi.org/10.1049/nbt2.12115.

McCarthy DJ, Campbell KR, Lun ATL, Willis QF (2017). “Scater: pre-processing, quality control, normalisation and visualisation of single-cell RNA-seq data in R.” _Bioinformatics_, *33*, 1179-1186. doi:10.1093/bioinformatics/btw777 https://doi.org/10.1093/bioinformatics/btw777.

OpenAI. (2025). ChatGPT. ChatGPT 5; OpenAI. https://chatgpt.com/.

R Core Team (2025). _R: A Language and Environment for Statistical Computing_. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/.

Shiny - File Upload. (2014, July 29). Shiny. https://shiny.posit.co/r/gallery/widgets/file-upload/

Shiny - Including HTML, text, and Markdown files. (2014, July 29). Shiny. https://shiny.posit.co/r/gallery/application-layout/including-html-text-and-markdown-files/

Shiny - Tabsets. (2014, July 30). Shiny. https://shiny.posit.co/r/gallery/application-layout/tabsets/

Wickham H (2016). ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. ISBN 978-3-319-24277-4, https://ggplot2.tidyverse.org.

Ziebell F (2024). _RNAseqQC: Quality Control for RNA-Seq Data_. doi:10.32614/CRAN.package.RNAseqQC https://doi.org/10.32614/CRAN.package.RNAseqQC, Rpackage version 0.2.1, https://CRAN.R-project.org/package=RNAseqQC.

This package was developed as part of an assessment for 2025 BCB410H: Applied Bioinformatics course at the University of Toronto, Toronto, CANADA. QCseq welcomes issues, enhancement requests, and other contributions. To submit an issue, use the GitHub issues.