Tools for analysis of Luminex immunoassay data: development of a robust pipeline and best practices recommendations

Date
2023-02
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
ENGLISH ABSTRACT: Background and Scope: Immunoassays can be used to detect and measure the concentration of many antigens in a variety of specimens for the diagnosis of diseases, and for the detection of microbes and various illegal substances. In addition, they can be used to monitor and study processes, or differentiate infection, latent or active disease in a patient's immune response after infection with a pathogen by measuring the presence of specific antigens. Recent advances in the instrumentation include multiplex immuno-assays, e.g., Luminex 200 or Luminex MAGPIX®, powered by Luminex xMAP® technology. Analysis of data produced by the multiplex immunoassays is complex and current analytical approaches are highly subjective. Currently, there is no standard, robust approach for data pre-processing of multiplex immunoassay data. To overcome this knowledge gap, I developed a robust and standardized data pre-processing pipeline for multiplex immunoassay data to ensure reproducible data. In addition, provide recommendations for best practices of Luminex data generation and data pre-processing to ensure reproducible science by amending existing laboratory standard operating procedure (SOPs) and developed data pre-processing SOPs’ for the Stellenbosch University Bioinformatics Research Group. Design and Approach: I evaluated current laboratory processes, analysed existing Luminex data and drafted best practices recommendations that will ensure reliable input data for the pipeline. I implemented programmatic steps for data management, quality control, and pre-processing to provide high quality data for analyses. A variety of data preprocessing approaches were investigated, and a set of robust standardized options are provided to the user together with appropriate bioinformatics tools in the newly developed pipeline. Results: For robust and reproducible data generation and data preparation, standardization of manual laboratory best practices procedures are recommended and implemented to ensure standard file name convention, file format, file structure and standard plate layout. In addition, the development of a robust automated standardized data pre-processing pipeline using algorithms in R, freely available software will reduce variability and error introduced by humans. In addition, to ensure reliability and reproducibility of the results generated using this pipeline, the pipeline records metadata such as parameter settings, program, and package versions in the output. The methods were validated using existing de-identified Stellenbosch University https://scholar.sun.ac.za iv data sets from the Stellenbosch University Immunology Research Group. In the future the pipeline can be applied to newly generated data from a variety of immunological studies. Conclusions: This pipeline will standardize and speed up data pre-processing, as well as provide consistent and reproducible results with any complex analyses. Furthermore, provide the bioinformatician or statistician with a rapid means to pre-process Luminex data for subsequent analysis. The framework developed here can be easily applied to other data analysis projects from different biomedical fields.
AFRIKAANS OPSOMMING: Agtergrond en omvang: Immunotoetsings kan gebruik word om die konsentrasie van baie antigene in 'n verskeidenheid monsters op te spoor en te meet vir die diagnose van siektes, en vir die opsporing van mikrobes en verskeie onwettige stowwe. Daarbenewens kan hulle gebruik word om prosesse te monitor en te bestudeer, of om infeksie, latente of aktiewe siekte in 'n pasiënt se immuunrespons na infeksie met 'n patogeen te onderskei deur die teenwoordigheid van spesifieke antigene te meet. Onlangse vooruitgang in die instrumentasie sluit multipleks-immuno-toetse in, bv. Luminex 200 of Luminex MAGPIX®, aangedryf deur Luminex xMAP-tegnologie. Ontleding van data wat deur die multipleksimmunotoetse geproduseer word, is kompleks en huidige analitiese benaderings is hoogs subjektief. Tans is daar geen standaard, robuuste benadering vir data voorverwerking van multipleks immunoassay nie. Om hierdie kennisgaping te oorkom, het ek 'n robuuste en gestandaardiseerde data-voorverwerkingspyplyn vir multipleks-immunotoetsdata ontwikkel om reproduceerbare data te verseker. Dit verskaf ook aanbevelings vir beste praktyke van Luminex-datagenerering en datavoorverwerking om reproduseerbare wetenskap te verseker deur bestaande laboratorium-standaard-operasionele-prosedure (SOPs) te wysig en data-voorverwerking-SOPs vir die Universiteit Stellenbosch Bioinformatika Navorsingsgroep te wysig. Ontwerp en Benadering: Ek het huidige laboratoriumprosesse geëvalueer, bestaande Luminex-data ontleed en beste praktyke-aanbevelings opgestel wat betroubare insetdata vir die pyplyn sal verseker. Ek het programmatiese stappe vir databestuur, kwaliteitbeheer en voorafverwerking geïmplementeer om data van hoë gehalte vir ontledings te verskaf. 'n Verskeidenheid datavoorverwerkingsbenaderings is ondersoek, en 'n stel robuuste gestandaardiseerde opsies word aan die gebruiker verskaf saam met toepaslike bioinformatika-instrumente in die nuut ontwikkelde pyplyn. Resultate: Vir robuuste en reproduseerbare datagenerering en datavoorbereiding, word standaardisering van prosedures vir beste praktyke in die laboratorium aanbeveel en geïmplementeer om standaard lêernaamkonvensie, lêerformaat, lêerstruktuur en standaardplaatuitleg te verseker. Daarbenewens het die ontwikkeling van 'n robuuste outomatiese gestandaardiseerde data voorafverwerking pyplyn deur gebruik te maak van algoritmes in R, vrylik beskikbare sagteware sal veranderlikheid en foute wat deur mense bekendgestel word, verminder. Bykomend tot betroubaarheid en reproduseerbaarheid van die resultate wat met hierdie pyplyn gegenereer word, te verseker, teken die pyplyn Stellenbosch University https://scholar.sun.ac.za vi metadata soos parameterinstellings, program- en pakketweergawes in die resultate aan. Die metodes is bekragtig deur gebruik te maak van bestaande gede-geïdentifiseerde datastelle van die Universiteit Stellenbosch Immunologie Navorsingsgroep. In die toekoms kan die pyplyn toegepas word op nuutgegenereerde data uit 'n verskeidenheid immunologiese studies. Gevolgtrekkings: Hierdie pyplyn sal datavoorverwerking standaardiseer en bespoedig, asook konsekwente en reproduseerbare resultate met enige komplekse ontledings verskaf. Voorsien ook die bioinformatikus of statistikus van 'n vinnige manier om Luminexdata vooraf te verwerk vir daaropvolgende analise. Die raamwerk wat hier ontwikkel is, kan maklik toegepas word op ander data-ontledingsprojekte uit verskillende biomediese velde.
Description
Thesis (PhD)--Stellenbosch University, 2023.
Keywords
Citation