Thema der Dissertation:
Association Plots visualize cluster-specific genes from high-dimensional transcriptomics data Thema der Disputation:
Dimensionality reduction methods
Association Plots visualize cluster-specific genes from high-dimensional transcriptomics data Thema der Disputation:
Dimensionality reduction methods
Abstract: Dimensionality reduction methods allow to reduce the size of a data while retaining as much information as needed. In my talk I will focus on two linear methods for dimensionality reduction. First, I will present the basic concepts of principal component analysis (PCA), a method commonly used in transcriptomics data analysis. Second, I will focus on correspondence analysis, a method similar to PCA which however enables the joint embedding of variables and observations. Finally, I will introduce Association Plots, a newly developed method for identification of cluster-specific genes. Association Plots are derived from correspondence analysis and allow for visualization of gene-cluster associations in high-dimensional transcriptomics data.
Zeit & Ort
04.07.2022 | 14:00
Ort: Seminarraum 049
(Fachbereich Mathematik und Informatik, Takustr. 9, 14195 Berlin)