Method For Automated Square Wave Electrochemical Sensing Data Pre-Processing

Abstract Square Wave Voltammetry (SWV) is a common technique used in biosensing. This method is used with modified electrodes which can detect biomarkers associated with many illnesses. However, large volumes of data are collected, creating a bottleneck in data analysis in research. A significant amount of time is wasted by researchers on manually analyzing and preparing data for presentation. Given that SWV biosensing is a very common technique, this application developed using Python 3.10.1, which can be used as a standalone application in Windows or can be accessed using a Python interpreter on other Operating Systems, can greatly streamline research in SWV Biosensing. GitHub Repository Link: https://github.com/agosh-saini/Ideation_Automation Introduction Square Wave Voltammetry is a common analysis technique with use common in diagnostics, kinetics studies, and biological analysis [1]. Square wave voltammetry consists of a square wave voltage pulse which then allows the user to analyze the current [1]. The software creates a simple GUI for the user to input variables that help determine the metadata section in the file and the general structure of the data. This GUI creates a system that can adjust to most systems used for SWV biosensing. The software scans the relevant sections for peaks or valleys that meet the predetermined thresholds, marking them on a graph and in a separate comma-separated file. SWV signals are generated due to electrochemical surface interactions[2], and hence are present in known ranges, allowing for the automation of data analysis. This data can then be used for analysis and experimental design. The software was developed using common packages, such as SciPy, NumPy, and Pandas. The list of environmental modules can be found in the requirements file in the repository. The use of Python and the function-focused design allows users to easily add additional functionality without significant modification to the existing program. The system described below is used by the IDEATION lab at the University of Waterloo in research. Papers in the space focused on the integration of Artificial Intelligence (AI) tools and advanced denoising functions, however, there was a clear gap in the space that this program tries to fill by creating a simple and easy-to-use tool that specifically automates data detection and graphing. Such a system is essential for creating large sets of varied data required for a robust AI model. Such a software tool would also be helpful in training Machine Learning (ML) models with high-quality labeled data. The significant amount of varied data required for creating a robust ML model is one of the key concerns of researchers in the field [3]. Common software that has similar functionality and is either free or easily accessible to most researchers include: Excel Visual Basic for Applications (VBA) Excel is commonly used for manually analyzing data by researchers. VBA allows for automation of common repetitive tasks such as report generation. There is a learning curve for individuals who do not have a programming background. Pandas Profiling This is a Python module that is optimized for visualizing datasets and generating reports with given data, including summaries of statistics that may be relevant to the user [4]. The customization is limited and it is not focused on SWV voltammetry. Paid data analysis tools (Polymer Search, Zoho Analytics, Prism, etc.) Powerful tools that can generate reports, and analyze data, and some of the tools have AI integrations to help with the process of analysis, however, such tools. These tools are often paid, limiting access, and are often geared towards general data analysis and/or data analysis for Machine Learning or Business Development applications. Implementation and Architecture Main.py (Python 3.10.1): Custom Modules: Ideation_ec_automation Standard and third-party Libraries: OS (Standard Library) ast (Standard Library) PySimpleGUI (Third Party – 5.0.6) [5] Structure: UI is created using PySimpleGU. The UI is simple and consists of checkboxes, text input elements, folder selection elements, and lists. The UI is used to collect all the information which is stored in a dictionary format. The user inputs are used as function inputs for functions from the custom module. Ideation_ec_automation.py (Python 3.10.1): Standard and third-party Libraries: Matplotlib (Third party – 3.9.1) [6] NumPy (Third party – 3.1.0) [7] Pandas (Third party – 2.2.2) [8] SciPy (Third party – 1.14.1) [9] OS (Standard Library) Glob (Standard Library) Shutil (Standard Library) Itertools (Standard Library) Datetime (Standard Library) Re (Standard Library) Typing (Standard Library) Structure: The class “ideation_ec_automation” consists of three sections. File processing functions (convert_csv, get_header_line, format_file_to_csv, create_df). These functions help create a standard file for further analysis using Pandas, NumPy, and SciPy Data analysis methods (plot_res, get_max_min, get_baseline, get_peaks, insert_signals) are focused on determining the peaks and valleys in the data, and marking these values correctly on generated figures and Data storage methods (convert_deliminator, delete_temp_files) organize the data for the user in relevant folders and remove unneeded files. Architecture Decisions: Modularity is the key goal during the development of the program. Each system has different needs and the user should have the ability to modify the program to meet their needs. Furthermore, there is a focus on using pre-existing libraries for data handling due to the high levels of optimization in many of the common third-party libraries such as NumPy, leading to faster execution time for large numbers of files or large file sizes. The use of PySimpleGUI further highlights the focus on simplicity. The Graphical User Interface (GUI) consists only of the parameters that are necessary for researchers as to have a low barrier to learning to use the software. This decision allows researchers to focus primarily on setting up and conducting tests, without the need to learn to program or complex GUIs found with alternative free solutions such as Pandas Profiling or Excel VBA. Reuse Potential SWV is a very common analysis technique with a growing interest in using Machine Learning (ML) to assist in analysis [10]. Given that many labs conduct research with novel antigens/analytes, high-quality labeled data is essential for ML training. The problem of data analysis being a research bottleneck was determined after conversations with researchers within and outside of the lab group. Alternative solutions for SWV analysis have high skill or payment barriers which force researchers to conduct analysis manually. This program should significantly increase the productivity of individual researchers. A similar problem is present in adjacent research topics such as material property analysis. The modular nature of the program allows the user to reuse the software for said adjacent research topics with very few modifications as the general principle of peak detection is extremely common. Funding Statement Agosh Saini has been funded by the Waterloo Institute of Nanotechnology and Velocity during the course of this project. Fatemeh Keyvani has been funded by the Ontario Graduate Scholarship and the Natural Science and Engineering Council of Canada during the course of this project. The Centre for Bioengineering & Biotechnology (CBB)-Seed Funding received by Mahla Poudineh. Competing Interests The authors declare that they have no competing interests. Additional Information on the Development Environment Operating system Windows 11 Programming language Python - Version 3.10.1 Windows Command Line Script (Batch Script) - Tested on Windows 11 Dependencies contourpy==1.2.1 cycler==0.12.1 fonttools==4.53.1 kiwisolver==1.4.5 matplotlib==3.9.2 numpy==2.1.0 packaging==24.1 pandas==2.2.2 pillow==10.4.0 pyasn1==0.6.0 pyparsing==3.1.4 PySimpleGUI==5.0.6 python-dateutil==2.9.0.post0 pytz==2024.1 rsa==4.9 scipy==1.14.1 six==1.16.0 tzdata==2024.1 References [1] Chen, A and Shah, B 2013 Electrochemical sensing and biosensing based on square wave voltammetry. Anal. Methods, 5(9): 2158–2173. DOI: https://doi.org/10.1039/C3AY40155C [2] Mirceski, V, Komorsky-Lovric, S, and Lovric, M 2007 Square-Wave Voltammetry: Theory and Application. Springer. URL https://doi.org/10.1007/978-3-540-73740-7 [3] Kang, M, Kim, D, Kim, J, Kim, N, and Lee, S 2024 Strategies to Enrich Electrochemical Sensing Data with Analytical Relevance for Machine Learning Applications: A Focused Review. Sensors, 24(12). DOI: https://doi.org/10.3390/s24123855 [4] Pandas Profilling Documentation. URL https://pypi.org/project/pandas-profiling/ [5] PySimpleGUI Documentation. URL https://docs.pysimplegui.com/en/latest/ [6] Matplotlib 3.9.2 documentation. URL https://matplotlib.org/stable/index.html [7] NumPy user guide. URL https://numpy.org/doc/stable/user/index.html#user [8] Pandas Documentation. URL https://pandas.pydata.org/docs/ [9] SciPy Documentation. URL https://docs.scipy.org/doc/scipy/ [10] Kayali, D, Shama, N A, Asir, S, and Dimililer, K 2023 Machine learning-based models for the qualitative classification of potassium ferrocyanide using electrochemical methods. The Journal of Supercomputing, 79(11): 12472–12491. DOI: https://doi.org/10.1007/s11227-023-05137-y

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.