BioVar: an online biological variation analysis tool

S Korkmaz, G Zararsız, D Göksülük, M Senes, C Sönmez, D Yücel

Abstract

Objectives: Biological variation (BV) analysis of laboratory tests gets increased attention due to its practical applications. These applications include correct interpretation of laboratory tests, the decision on the availability of reference intervals, contributions to clinical decision-making. It is critical to derive the BV information accurately and reliably. Another crucial step is to perform the statistical analysis of the BV data. Although there are updated and comprehensive guidelines, there is no reliable and comprehensive tool to perform statistical analysis of BV data. Methods: We presented BioVar, an online tool for statistical analysis of the BV data based on available and updated guidelines. Results: This tool can be used (i) to detect outliers, (ii) to control normality assumption, (iii) to check steady-state condition, (iv) to test homogeneity assumptions, (v) to perform subset analysis for genders, (vi) to perform analysis of variance to estimate components of variation and (vii) to identify analytical performance specifications of laboratory tests. Moreover, plots can be created at each step of outlier detection to inspect outliers and compare gender groups visually. An automatic report can be generated and downloaded. Conclusion: The tool is freely available through turcosa.shinyapps.io/biovar/, and source code is available on the Github: github.com/selcukorkmaz/BioVar.