Available in this repository:
-
Data preprocessing →
scRNA_analysis_QC.py
Create QC plots to preprocess data and select highly variable genes.
-
Dimension Reduction and Annotation →
scRNA_analysis_DR.py
Create plots for biological insight.
Data set contained 14,783 cells and 17,374 genes.
UMAP of cells annotated by predicted cell type
Cell type description:
- B cells memory: Long-lived lymphocytes that retain the specific antibody information from a previous infection, enabling faster and stronger antibody response upon re-exposure to the same pathogen.
- B cells naive: Mature but inexperienced B lymphocytes that have not yet encountered a specific antigen.
- Erythroid-like and erythroid precursor cells: Hematopoietic cells in the bone marrow that are on the path to becoming mature, oxygen-carrying red blood cells (erythrocytes).
- Gamma delta T cells: Unique subset of T cells that patrol epithelial and mucosal surfaces, acting as a bridge between the innate and adaptive immune systems.
- NK cells (Natural Killer cells): Innate lymphocytes that kill host cells lacking 'self' markers (MHC class I), making them crucial for early defense against viruses and cancer.
- Neutrophils: Most abundant white blood cell, for phagocytosing and killing invading bacteria.
- Paneth cells: Specialized epithelial cells located in the small intestine and secrete antimicrobial peptides to maintain the gut barrier.
- Pericytes: Mural cells embedded in the basement membrane of capillaries and venules and help with regulating blood-brain barrier integrity.
- Plasma cells: Fully differentiated B cell progeny that are dedicated antibody factories, secreting large amounts of a single type of antibody.
- Platelets: Small, anucleate cell fragments important for blood clotting and wound repair.
- T cells: Key lymphocytes of the adaptive immune system that differentiate into cytotoxic killers or helper cells.
- Thymocytes: Developing lymphocytes found in the thymus that undergo selection processes to mature into functional, self-tolerant T cells.
This dataset is not consistent with a bone marrow sample. A bone marrow sample is the primary site of adult hematopoiesis, meaning it should contain majority hematopoietic stem/progenitors. This lineage is not present at all in the UMAP. Instead, the UMAP plot successfully displays expected lineages in a blood sample population:
- Myeloid Monocytes: Neutrophils and Macrophages
- Lymphoid: B-cells, T-cells, NK cells, Plasma Cells
- Erythroid
- Progenitors/Stem Cells
Upon further analysis, the presence of individual bone marrow markers where searched. These included $CD34, $PROM1, and $KIT. None of these were found. The lack of these classical progenitor markers strongly suggest that data is likely from peripheral blood mononuculear cells (PBMCs), blood immune cells, or tissue-resident immune cells. Another abnormal findings to suggest that the data is not from a bone marrow sample are the presence of non-hematopoietic cells such as paneth cells and pericytes.
Dotplot of marker genes
The dominance of immature populations such as hematopoietic stem cells and myeloid cells , are the hallmarks of an active hematopoietic organ like the bone marrow. Our data set shows T-cells to be the most abundant population, strongly suggesting these are not bone marrow samples. T-cells and NK cells are present in abundance in PBMCs.
Heatmap of marker genes
Our dataset shows high expression of markers like
Cell type count
The cell count bar plot reveals an overwhelming dominance of Gamma delta T-cells. Thus suggesting infected cell data.
Expression pattern of Interferon-Stimulated Genes (ISGs) and Interleukin (IL) receptors
High expressions of
The data is highly consistent with patients undergoing a significant infectious or inflammatory episode, most likely triggered by a viral pathogen, given the strong interferon-stimulated gene signature and the extensive activation and proliferation of the cytotoxic T and NK cell lineages.