An open-source coding basis, rapid and integrated method of bar chart visualization combined with multi-comparison analysis for significant differences

While the concept of multivariable analysis has been embraced for many decades, with the extensive implementation of high-throughput omics technologies, the revolution has been placed in the field to incorporate big data analysis including bioinformatics, statistics, machine learning, and deep learning. However, in reality, researchers have limited time and ability to process data in a very efficient and accurate way.

To address this issue and increase the workflow efficiency for scientific research, an open-source coding-based method of bar chart visualization combined with multi-comparison analysis for significant differences has been developed. The Python code reads data directly from a CSV table file with multivariable within multiple samples/ treatments, then generates a bar chart with color differences and different pattern fills in bars of variables, which improves the visual differentiation and data readability for both digital and black-white print reading. Also, standard deviation and significant differences based on the selected least significant difference (LSD) test or Turkey test are automatically marked on the top of each bar. Compared to the commonly used bar charts such as in MS Excel software combining significant differences analysis in R software, which is usually a time-consuming ‘Two-Step’ operation, this open-source coding basis bar chart method rapidly integrates both data visualization and statistical analysis results in seconds.

This open-source code aims to provide liberation for researchers from traditional, labor-intensive, and time-consuming manual data analysis. Two coding examples with annotations and completed user guidance are also provided for users to learn coding-based statistical data analysis and to help them utilize such techniques for their future research.