Detalle de: Response to the letter ‘testing the effectiveness of MyPROSLE in classifying patients with lupus nephritis’

Recently, a letter to the Editor entitled ‘Testing the Effectiveness of MyPROSLE in Classifying Patients with Lupus Nephritis’ has been submitted by Leventhal et al. to Briefings in Bioinformatics. In this letter, the authors test MyPROSLE, a web application we recently introduced [1], to characterizelupus patients from the molecular point of view. Leventhal et al. tested the application with independent datasets reporting that the software ‘did not perform sufficiently well to consider replacement of the standard kidney biopsy as a diagnostic procedure’. In this letter, we address in detail all the concerns described by Leventhal et al. First of all, we would like to thank the authors for their interest and evaluation of the web tool. Nevertheless, we want to remark that in this work we did not intend to provide software for the replacement of standard clinical diagnostic procedures, as they stated. In our manuscript, we present a scoring system to summarize the molecular portrait of each patient and machine learning models based on these features are one of the analyses used to demonstrate the utility of this scoring system. In this context, MyPROSLE was developed to apply this scoring system to gene expression datasets in order to predict clinical features based on transcriptomics data. The letter is focused on the performance of these models but, although transcriptomics profiles have emerged as a valuable resource for making new and significant discoveries in diagnosis, the integration of these profiles into clinical practice is still a distant goal. Consequently, our software was not designed to replace existing diagnostic approaches in the clinical setting but a system that can provide additional information for clinical decisions when sufficient quality RNA-Seq data is available. In the current scenario, it should be used for exploratory analysis and hypothesis generation. This concept is what we embodied in the final sentence of our original article: ‘Therefore, we set a precedent and an important advance in terms of personalized research’ (through the development of an analytical workflow) ‘oriented to a near future clinical practice within autoimmunity’.