Analysis of qRT-PCR Data to Identify the Most Stable Reference Gene Using gQuant

Equipment

1. For optimal performance, gQuant requires a system with a multi-core processor (Intel Core i5 or equivalent and above) and 8 GB RAM and above to efficiently handle computation. The tool is compatible with Windows, macOS, and Linux-based operating systems supporting Python 3.7 or higher. Additionally, using the Conda distribution with Jupyter Notebook (version 6.0 and above, optional) is recommended for an integrated development environment.

Software and datasets

Python software environment (≥3.6) (https://www.python.org/)

Anaconda distribution with Jupyter Notebook integrated development environment (≥6.0, optional) (https://www.anaconda.com/)

All data and code have been deposited to GitHub: https://github.com/ABHAYHBB/gQuant

1. Datasets can be prepared in different ways, and the scope of the protocol does not cover the processing of raw data. Generally, data should be cleaned, quality checked, and ready to be used.

2. Supported file formats for the tool can be, e.g., Excel files (.xlsx), comma-separated files (.csv), text files (.txt), and data frames (df).

3. After obtaining the qRT-PCR data, input source type does not matter, and the following requirements are the same:

• Columns of the dataset must represent the name of the gene/target and the row represents the sample.

• The name of the rows (samples) must be removed.

• For the best practice, avoid having spaces in the column names. It could cause unnecessary problems.

• If the dataset has missing values, it will not affect the result because it is carefully designed to handle it.

4. The datasets used here to show the example results are the same as what was used in the original article. These three datasets collectively validate gQuant across different molecular targets (mRNA vs. miRNA), sample types (cell lines vs. patient samples), and experimental constraints (small vs. large datasets, missing values). By demonstrating its robustness in diverse conditions, we confirm that gQuant can be reliably applied to various qRT-PCR-based normalization scenarios.

a. Dataset 1: Derived from research on OvCar-3 and PC-3 cancer cell lines, this dataset encompasses transcriptomic profiling of 84 genes pertinent to cell regulation and five conventional housekeeping genes (Figure 2). The data, accessible via GSE57888, serves as a benchmark for mRNA-based qRT-PCR analysis, offering validation for gQuant through non-normalized data. It contains un-normalized Ct values, where column names represent gene names, and rows correspond to samples without explicit row labels. Data also have undetermined or missing values. The file can be accessed using the link https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE57888.

Figure 2. Example of expression data in Excel file (.xlsx). The dataset shown in the figure is Dataset 1 as defined in this article.

b. Dataset 2: We chose the dataset GSE239868 with more detailed variables, such as the expression of 1,066 human miRNAs and n = 36 tracheal aspirate samples, for a comparative efficiency assessment of gQuant. This database included data for a comparative analysis on the management of missing values and allowed us to test the appropriateness of our method on miRNA background. This dataset’s unnormalized Ct values provide a strict testing environment. The file can be accessed using the link https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi.

c. Dataset 3: In an endeavor to construct a more robust analytical framework, we decided to explore the potential of miRNA-based normalizers as an alternative to the well-established mRNA-based normalizers. Dataset 3 is also a testing environment for small datasets. Given the nascent state of miRNA-based normalization data, we focused specifically on uEVs. A comprehensive literature review and Qiagen’s Human Urine Exosomes Focus miRCURY LNA Panel list were used to select the miRNA set used in this study for validation. This investigation led to the selection of miR-16-5p, miR-10b-5p, 30b-5p, and miR-30d-5p. Additionally, upon evaluating the expression levels of the Let7c cluster genes from another ongoing study, we observed high stability for let-7c-5p across both diseased and control samples, leading to its inclusion in subsequent investigations. Therefore, our preliminary study proceeded to focus on the selected miRNAs: let-7c-5p, miR-16-5p, miR-10b-5p, miR-30a-5p, and miR-30d-5p. Using qRT-PCR, we assessed their expression in uEVs samples of BCa (n = 9), PCa (n = 6), and control samples (n = 3). Among the evaluated miRNAs, let-7c-5p has emerged as the most promising candidate for normalizer functions. Subsequently, expression levels for let-7c-5p miRNAs in uEVs were quantified in an expanded sample set (n = 30) using qRT-PCR. Rigorous quality control measures, including primer efficiency and melt curve analysis, were implemented, and any data with Ct values over 38 were excluded. The stability of the miRNA expression was validated using gQuant, solidifying let-7c-5p's potential as a reliable normalizer. This dataset was uploaded to the gQuant tool GitHub repository (https://github.com/ABHAYHBB/gQuant) (Dataset S1. Dataset 3).

Procedure

A. Installing Python and Jupyter

Installing Python and Jupyter Notebook is not covered by the presented protocol; however, detailed guidance on how to install them can be found on their respective official website.

Python: https://www.python.org/

Jupyter Notebook: https://www.anaconda.com/

B. Installing the required dependencies for the tool (File S1)

The gQuant tool has a few dependencies that need to be installed.

1. Data preprocessing: These libraries are used for data manipulation and handling missing values.

Pandas: !pip install pandas

NumPy: !pip install numpy

2. gQuant tool computation: These libraries are used for statistical computations, scaling, and voting classifiers.

SciPy: !pip install scipy

Scikit-Learn: !pip install scikit-learn

3. Data visualization: These libraries are used for generating publication-ready interactive and static visualizations.

Bokeh: !pip install scikit-learn

Plotly: !pip install scikit-learn

C. Importing the qRT-PCR expression data

Code is given below for various supported file formats:

1. Spreadsheet (Excel):

import pandas as pd

df_excel = pd.read_excel('path_to_your_file.xlsx', sheet_name = 0, header = 0)

Parameter explanation: Change this parameter as your path and file name “'path_to_your_file.xlsx”, provide sheet name in this parameter “sheet_name = 0” (it can either be a number or name of the sheet), and the header parameter “header = 0” should be set to 0 as it reads the first row as a column name.

2. Text file (plain text):

df_text = pd.read_csv('path_to_your_file.txt',delimiter = '\t', header = 0)

Parameter explanation: Change this parameter as your path and file name “'path_to_your_file.txt” and the delimiter parameter “delimiter = '\t'” is used to set the delimiter sign, which is used in the text file.

3. Comma-separated (CSV) files:

df_csv = pd.read_csv('path_to_your_file.csv', header=0)

Parameter explanation: Change this parameter as your path and file name “path_to_your_file.txt” and the header parameter “header=0” is set to 0, assuming the first row has the column names.

D. Data preprocessing

1. For data preprocessing, the user needs to import the libraries that have been installed earlier in this section.

import pandas as pd

import numpy as np

2. Copy the function code “preprocessing.py” from the GitHub repository of the tool, paste it into the Jupyter notebook, and run the function (Figure 3). Figure 4 presents the dummy expression data with missing values (NAs).

Figure 3. Snapshot of the GitHub repository of gQuant showing preprocessing function code

Figure 4. Initial dummy expression data with missing values

3. To use the preprocessing function, use the function call syntax with the imported file name as “df”.

results, df1 = preprocessing(df)

4. To display the results, use the following syntax

print("Overall Results:", results)

to print the overall results showing the count ratio of missing values to available values of the expression data (Figure 5).

print("Updated File:", df1)

Figure 5. Output of the preprocessing function on dummy expression data as a data frame “result”, which uses an exclusion ratio threshold of 8:1 (available values to missing values)

This will print the initial samples of preprocessed data named “df1” (Figure 6).

Figure 6. Final data frame with filtered columns and data imputation for “Gene1”

5. The preprocessing function code given in the tool uses a default exclusion ratio threshold of 8:1 of available values to missing values. This ratio can be changed according to the user’s needs. To change this ratio, the user can modify the function code as follows:

df1=results[(results['Available_to_Missing_Ratio']>=8)|(results['Missing_Count'] == 0)]

6. To save this “df1” file in CSV format, use this syntax

df1.to_csv('df1.csv', index=True)

E. gQuant tool

1. To use the gQuant tool, the user needs to import the necessary libraries that were installed earlier in this section.

import pandas as pd

import numpy as np

from scipy.stats import gmean

from sklearn.neighbors import KernelDensity

from sklearn.preprocessing import MinMaxScaler

2. Copy the function code “tool.py” from the GitHub repository, paste it into the Jupyter notebook, and run the function (Figure 7).

Figure 7. Snapshot of the GitHub repository of gQuant, showing the tool function code

3. After running the tool function “tool.py”, use the syntax given below to call the function passing the data frame “df1”, which was preprocessed and downloaded from the preprocessing function.

ranking_df = gquant(df1)

After calling the function, the output will be in table format (Figure 8). If there are multiple genes, all the rows of the table may not be visible. To see the full ranking table, use the next instruction.

Figure 8. Example output of gQuant tool

4. To save the ranking table that was generated by the “tool.py” function, use the following syntax:

ranking_df.to_csv('ranking_table.csv', index=False)

F. Data visualization

1. For data visualization (boxplot and KDE plot), import the libraries that were installed earlier in this section.

import plotly.express as px

import plotly.graph_objects as go

import pandas as pd

import numpy as np

from scipy.stats import norm

from bokeh.plotting import figure, show

from bokeh.io import output_notebook

from bokeh.models import LabelSet, ColumnDataSourceimport

2. Copy the function code “visualization.py” from the GitHub repository, paste it on the Jupyter notebook, and run the function (Figure 9).

Figure 9. Snapshot of the GitHub repository of gQuant, showing the visualization function code

3. To create a boxplot and KDE plot, the user needs to call the boxplot function as given below:

create_box_plot(df1)

create_kde_plot(df1)

After calling these visualization functions, the user will see the output as two different plots based on the dataset. The high-quality .png images will be saved on the local file system and named “kde_plot.png” and “box_plot.png”. The boxplot in Figure 10 shows the expression differences and distributions of Dataset 3.

Figure 10. Expression levels presented in the form of boxplot using Dataset 3

The KDE plot in Figure 11 demonstrates the density of the data points applied on Dataset 3.

Figure 11. Kernel density estimation level plot using Dataset 3

Data analysis

The step-by-step guide “Tutorial Document” is provided in File S1, which is user-friendly even for beginners in Python programming. This also provides options for exporting ready-to-publish figures and tables.

Validation of protocol

This protocol or parts of it has been used and validated in the following research article:

• Pathak et al. [20]. Development of a robust and generalizable algorithm "gQuant" for accurate normalizer gene selection in qRT-PCR analysis. Scientific Reports. (Figure 2 and Tables 1, 2, 3, and 4).

General notes and troubleshooting

General notes

1. The gQuant tool is created using Python programming environment. Users should have a basic understanding of the Python environment and any integrated development environment (preferred: Jupyter notebook).

2. Before using the gQuant tool, make sure you have the dataset in the right format.

3. A Python environment setup is necessary to utilize the tool efficiently. The user can search in the gQuant GitHub repository for dependencies. Make sure that you have all possible libraries installed and imported in the correct version as specified.

4. Be aware of the compatibility of the environment; updating the software and libraries reduces error occurrences when running the algorithm code.

5. The visualizations created by the functions given here create high-quality pictures. Users can also change the format with small changes in the code. This type of information is not included in this protocol.

6. For optimal stability ranking, we recommend using at least three technical replicates per gene and three biological replicates per group, although gQuant can process datasets of any size and gene count.

7. To mitigate sample size constraints, publicly available datasets can be integrated into the analysis. However, users should carefully consider the following criteria to ensure data compatibility and reliability:

• Ensuring comparable experimental conditions.

• Checking data normalization and quality control steps.

• Addressing potential batch effects that could introduce bias.

• By supplementing small-scale experimental data with validated public datasets, researchers could improve the robustness of stability rankings in cases where the number of experimental replicates is limited.

8. gQuant is compatible with Ct data obtained from both SYBR Green and TaqMan probe-based qRT-PCR assays. The tool operates on unnormalized Ct values using robust statistical methods that are independent of the probe chemistry. While experimental validation has been primarily conducted with SYBR Green data, the underlying statistical framework ensures consistent performance across both detection methods.

Troubleshooting

Problem 1: Error in missing value imputations.

Possible cause: A dataset containing non-numeric entities or incorrectly formatted values.

Solution: Make sure that all the entries are in numerical format; if not, make sure to format them correctly.

Problem 2: Incorrect gene rank output.

Possible cause: Improper preprocessing of data.

Solution: Verify that data holds the desired format of information and has been preprocessed using the preprocessing tool. Also, make sure that column-to-gene structure is being maintained in the process.

Problem 3: Failed to generate visualizations (boxplot/KDE plot).

Possible cause: Missing libraries or incorrect function.

Solution: Ensure that all the required libraries were installed and imported in the correct way. Make sure that the correct data is being passed to visualization functions.

Problem 4: Tool crashes while performing gene ranking of large amounts of expression data.

Possible cause: Insufficient memory/computational resources.

Solution: Optimize by removing duplicates (if any). Use a high-end machine with good computational resources.

Problem 5: No output or empty ranking table.

Possible cause: Dataset filtering during preprocessing due to high missing value ratios.

Solution: Adjust to the appropriate threshold. Make sure that columns are not entirely missing.

Problem 6: Handling Ct variability and setting thresholds in gQuant.

Possible cause: Ct values in qRT-PCR datasets can be influenced by various experimental factors, such as sample quality, RNA extraction methods, qPCR platforms, fluorophores, and primer efficiencies. Additionally, unnormalized Ct values require users to define appropriate thresholds for data processing, which may introduce inconsistencies if not properly managed.

Solution: To address this, we provide the best practices for users to enhance data comparability and improve analysis consistency when using gQuant. Users should follow these guidelines:

• Use the same qPCR instrument and reagents for all samples to reduce inter-run variability.

• Maintain consistent RNA extraction and reverse transcription protocols to prevent batch effects.

• Verify primer efficiency and specificity through melt curve analysis and standard curve generation.

• Avoid including Ct values from degraded or low-quality RNA samples, as these can introduce bias in the normalization process.

• gQuant automatically imputes missing Ct values if their presence is below a user-defined threshold. However, users should manually inspect missing values to determine if imputation is appropriate or if certain genes should be excluded. The default gQuant setting removes genes where the available-to-missing value ratio is below 8:1. Users can adjust this threshold based on dataset size and experimental constraints.

• Identify and remove extreme Ct values that could arise from technical errors.

• While gQuant operates on raw Ct values, users should ensure that their dataset does not include systematic biases.

• Exclude Ct values above 38, as these indicate low template presence and unreliable amplification.

Supplementary information

The following supporting information can be downloaded here:

1. Dataset S1. Dataset 3

2. File S1. Tutorial Document (Step-by-step guideline for using the gQuant tool)

Terminology clarification

• Candidate reference sequence: Refers to any target or miRNA being analyzed with RT-PCR.

• Candidate reference sequence, reference gene, normalizer: These terms are used interchangeably throughout the manuscript.

• Most stable reference gene: Specifically used when discussing “candidate reference sequences” that are stably expressed across samples.

• mRNA-based normalization: Using genes as reference targets.

• miRNA-based normalization: May involve intergenic or non-coding RNA sequences.

• Missing values (NA): Refers to the absence of data in a dataset, represented as NaN, null, or NA.

• Multiple statistical methods: Refers to various techniques used to analyze data, such as SD, CV, KDE, and GM to draw insights from the data.

• Python: A high-level programming language widely used in data science, AI, and machine learning for its simplicity and extensive libraries.

• Jupyter: An open-source web application for creating and sharing documents that contain live code, equations, visualizations, and narrative text, commonly used for data analysis and scientific computing.

• Library: A collection of pre-written code in Python or other languages that provides specific functionality, such as NumPy, Pandas, and Matplotlib for data manipulation and analysis.

Acknowledgments

Conceptualization, Abhay Kumar Pathak, Garima Jain.; Investigation, Abhay Kumar Pathak, Garima Jain.; Writing—Original Draft, Abhay Kumar Pathak, Sukhad Kural, Shweta Singh, Garima Jain.; Writing—Review & Editing, Abhay Kumar Pathak, Garima Jain; Funding acquisition, Garima Jain.; Supervision, Garima Jain, Lalit Kumar, Manjari Gupta.

The authors would like to extend their gratitude to the Institute of Eminence (IoE), Banaras Hindu University for their financial support through IoE seed grant; they are also thankful to ACG JanCare fund for their financial support. We acknowledge the MIRNOW LLP for providing a Junior Research Fellowship (JRF) and Banaras Hindu University for a Non-Net fellowship.

This tool was published as an original article in Scientific Reports (Pathak et al. [20], DOI: 10.1038/s41598-024-66770-y).

Competing interests

The authors declare no conflicts of interest.

Ethical considerations

Dataset 1 and Dataset 2 were obtained from GEO (Gene Expression Omnibus), a public dataset. For the samples considered as Dataset 3, informed consent for participation was obtained from all individuals. Study protocols and samples were approved by the Human Ethical Committee, Institute of Science, BHU (Registration number ECR/226/indt/UP/2014/RR-22) and the Institutional Ethics Committee, Institute of Medical Sciences, BHU (Registration number ECR/526/inst/UP/2014/RR-20), confirming that all experiments were performed in accordance with relevant guidelines and regulations.

References

1. Ye, J., Xu, M., Tian, X., Cai, S. and Zeng, S. (2019). Research advances in the detection of miRNA. J Pharm Anal. 9(4): 217–226. https://doi.org/10.1016/j.jpha.2019.05.004
2. Harshitha, R. and Arunraj, D. R. (2021). Real‐time quantitative PCR: A tool for absolute and relative quantification. Biochem Mol Biol Educ. 49(5): 800–812. https://doi.org/10.1002/bmb.21552
3. Smith, T. A. D., AbdelKarem, O. A., Irlam-Jones, J. J., Lane, B., Valentine, H., Bibby, B. A. S., Denley, H., Choudhury, A. and West, C. M. L. (2020). Selection of endogenous control genes for normalising gene expression data derived from formalin-fixed paraffin-embedded tumour tissue. Sci Rep. 10(1): 17258. https://doi.org/10.1038/s41598-020-74380-7
4. Suzuki, T., Higgins, P. J. and Crawford, D. R. (2000). Control Selection for RNA Quantitation. BioTechniques. 29(2): 332–337. https://doi.org/10.2144/00292rv02
5. Shen, Y., Li, Y., Ye, F., Wang, F., Wan, X., Lu, W. and Xie, X. (2011). Identification of miR-23a as a novel microRNA normalizer for relative quantification in human uterine cervical tissues. Exp Mol Med. 43(6): 358–366. https://doi.org/10.3858/emm.2011.43.6.039
6. Veryaskina, Y. A., Titov, S. E. and Zhimulev, I. F. (2022). Reference Genes for qPCR-Based miRNA Expression Profiling in 14 Human Tissues. Med Princ Pract. 31(4): 322–332. https://doi.org/10.1159/000524283
7. Jacobsen, K. S., Nielsen, K. O., Winther, T. N., Glebe, D., Pociot, F. and Hogh, B. (2016). Identification of valid reference genes for microRNA expression studies in a hepatitis B virus replicating liver cell line. BMC Res Notes. 9(1): 38. https://doi.org/10.1186/s13104-016-1848-2
8. Bhatia, A., Upadhyay, A. K. and Sharma, S. (2023). miRNAs are now starring in “No Time to Die: Overcoming the chemoresistance in cancer.” IUBMB Life. 75(3): 238–256. https://doi.org/10.1002/iub.2652
9. Condrat, C. E., Thompson, D. C., Barbu, M. G., Bugnar, O. L., Boboc, A., Cretoiu, D., Suciu, N., Cretoiu, S. M. and Voinea, S. C. (2020). miRNAs as Biomarkers in Disease: Latest Findings Regarding Their Role in Diagnosis and Prognosis. Cells. 9(2): Article 2. https://doi.org/10.3390/cells9020276
10. Jahromi, F. N. A., Dowran, R. and Jafari, R. (2023). Recent advances in the roles of exosomal microRNAs (exomiRs) in hematologic neoplasms: pathogenesis, diagnosis, and treatment. Cell Commun Signal. 21(1): 88. https://doi.org/10.1186/s12964-023-01102-7
11. Piao, X.-M., Cha, E.-J., Yun, S. J. and Kim, W.-J. (2021). Role of Exosomal miRNA in Bladder Cancer: A Promising Liquid Biopsy Biomarker. Int J Mol Sci. 22(4): Article 4. https://doi.org/10.3390/ijms22041713
12. Sewer, A., Gubian, S., Kogel, U., Veljkovic, E., Han, W., Hengstermann, A., Peitsch, M. C. and Hoeng, J. (2014). Assessment of a novel multi-array normalization method based on spike-in control probes suitable for microRNA datasets with global decreases in expression. BMC Res Notes. 7(1): 302. https://doi.org/10.1186/1756-0500-7-302
13. Lekchnov, E. A., Amelina, E. V., Bryzgunova, O. E., Zaporozhchenko, I. A., Konoshenko, M. Y., Yarmoschuk, S. V., Murashov, I. S., Pashkovskaya, O. A., Gorizkii, A. M., Zheravin, A. A., et al. (2018). Searching for the Novel Specific Predictors of Prostate Cancer in Urine: The Analysis of 84 miRNA Expression. Int J Mol Sci. 19(12): Article 12. https://doi.org/10.3390/ijms19124088
14. Jain, G., Das, P., Ranjan, P., Neha, Valderrama, F. and Cieza-Borrella, C. (2023). Urinary extracellular vesicles miRNA—A new era of prostate cancer biomarkers. Front Genet. 14. https://doi.org/10.3389/fgene.2023.1065757
15. Pfaffl, M. W., Tichopad, A., Prgomet, C. and Neuvians, T. P. (2004). Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper – Excel-based tool using pair-wise correlations. Biotechnol Lett. 26(6): 509–515. https://doi.org/10.1023/B:BILE.0000019559.84305.47
16. Livak, K. J. and Schmittgen, T. D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 25(4): 402–408. https://doi.org/10.1006/meth.2001.1262
17. Köhsler, M., Leitsch, D., Müller, N. and Walochnik, J. (2020). Validation of reference genes for the normalization of RT-qPCR gene expression in Acanthamoeba spp. Sci Rep. 10(1): 10362. https://doi.org/10.1038/s41598-020-67035-0
18. St-Pierre, J., Grégoire, J.-C. and Vaillancourt, C. (2017). A simple method to assess group difference in RT-qPCR reference gene selection using GeNorm: The case of the placental sex. Sci Rep. 7(1): 16923. https://doi.org/10.1038/s41598-017-16916-y
19. Hoja-Łukowicz, D., Maciążek, D., Kościelniak, P. and Janik, M. E. (2022). Innovative GenExpA software for selecting suitable reference genes for reliable normalization of gene expression in melanoma. Sci Rep. 12(1): 3331. https://doi.org/10.1038/s41598-022-07257-6
20. Pathak, A. K., Kural, S., Singh, S., Kumar, L., Yadav, M., Gupta, M., Das, P. and Jain, G. (2024). Development of a robust and generalizable algorithm. Sci Rep. 14(1): 18774. https://doi.org/10.1038/s41598-024-66770-y