The single cell mRNA sequencing techniques now allow high-throughput and detailed transcriptomic analysis of single cells. Although the tissue type or even the organ from which the sample was taken is known, the cell type of the individual cells is not known. Even though all the cells in the sample come from the same tissue type, this doesn’t mean that they all have the same gene expression. Variations between individual cells can exist for functional reasons or be associated with disease and have important functional implications, affecting the tissues and organs of an organism. Thus valid and easy cell annotation in a tissue from scRNA-seq data is important. Workflows have been proposed with the main parts being pre-processing and structure identification to aid in scRNA-seq analysis. These include, but are not limited to, cell
annotation.

This thesis focus is to implement a web application which will help users to explore, analyse, visualize and annotate cells from scRNA-seq datasets by using multiple prior knowledge databases. For the analysis a workflow suggested by the bibliography will be used, together with a methodology which annotates single cells using prior knowledge sources. The application will be written in the Python programming language and will re-use the existing ecosystem of libraries for the scRNA-seq dataset, visualization and annotation of cells. A DevOps approach will be followed and will be able to run in Docker containers in order to simplify the installation and improve reproducibility.

As a result this application will offer to the user an easy way to select cell types of a tissue type or organ from one or more prior knowledge sources, to parameterize and execute the automatic analysis. During the automatic analysis, it will filter, enrich and annotate a cell type to each cell in the dataset. To help the user to better judge the quality of the analysis, it will visualize each step with figures and other information.