Thumbnails:
List:
Year:
Category:
Session:
Poster:
Getting poster data...
Gabor Beke, Milan Hucko, Lubos Klucar, Dana Cholujova, Jana Jakubikova (1Institute of Molecular Biology, Slovak Academy of Sciences, Bratislava, Slovakia; Cancer Research Institute, Biomedical Research Center, Slovak Academy of Sciences, Bratislava, Slovakia)
Mass cytometry (CyTOF) is a powerful technique for studying complex systems like the immune system. It generates high-throughput, high-dimensional data, the analysis of which requires advanced clustering or dimensionality reduction algorithms such as SPADE or viSNE. These workflows are resource-intensive and typically must be re-run to incorporate new samples. Additionally, annotating SPADE trees is time-consuming and demands extensive immunological knowledge. Visualize and annotate SPADE trees, we developed SPADEview — an interactive web portal built in R using packages such as Shiny, igraph, and others. Identified cell populations were used to train several machine learning models, including Neural Networks, Random Forests, and XGBoost (gradient-boosted trees), enabling rapid analysis of new samples without the need to re-run SPADE. To visually compare the results, we performed a t-SNE analysis and coloured the cells based on both the SPADE results and the predictions made by our models. This work was supported by grants APVV-19-0212, APVV-20-0183, APVV-23-0482.http://www.imb.savba.sk/index.php?id=departments&lang=en&labID=15