Data treatment methods for real-time colorimetric loop-mediated isothermal amplification reactions

With the SARS-CoV-2 pandemic and the need for affordable and rapid mass testing, colorimetric isothermal amplification reactions such as Loop-Mediated Isothermal Amplification (LAMP) are quickly rising in importance. The technique generates data that is similar to quantitative Polymerase Chain Reaction (qPCR), but instead of an endpoint color visualization, it is possible to construct a signal over a time curve. As the number of works using time-course analysis of isothermal reactions increases, there is a need to analyze data and standardize their related treatments quantitatively. Here, we take a step forward toward this goal by evaluating different available data treatments (curve models) for amplification curves, which allows for a cycle threshold-like parameter extraction. In this study, we uncover evidence of a double sigmoid equation as the most adequate model to describe amplification data from our remote diagnostics system and discuss possibilities for similar setups. We also demonstrate the use of multimodal Gompertz regression models. Thus, this work provides advances toward standardized and unbiased data reporting of Reverse Transcription (RT) LAMP reactions, which may facilitate and quicken assay interpretation, potentially enabling the application of machine learning techniques for further optimization and classification.