Welcome To Lymphoma EcoTyper

EcoTyper is a framework for the systematic identification of cell states and cellular communities (ecotypes) from bulk, single-cell and spatially-resolved gene expression data. This website allows users to explore features of the cell states and cellular communities identified in diffuse large B cell lymphoma (DLBCL) samples, as well as recover cell states and ecotypes in user provided expression data.

Website features:

  • Interact with, visualize, and explore EcoTyper results.
  • Download data associated with the EcoTyper publication.
  • Run EcoTyper to recover cell states and ecotypes on user provided bulk or scRNA-seq expression data.

The website does not allow users to discover new cell types and ecotypes in their own data, or perform cell state and ecotype recovery in spatial transcriptomics data. These functionalities are provided in the EcoTyper source code available in the Download section of this website and on the GitHub page of EcoTyper.

Explore Cell State Expression Signatures

  • Select a gene list and cell type.
  • This will display the expression signatures associated with all cell type specific states.
    • Hover or click and drag on the heatmap to explore genes and zoom.
    • To save time, the number of genes displayed per state is limited to 200. Download the full expression data to view all genes.
  • Core gene list is generated by CIBERSORTx in-silico purification. The extended gene lists is generated from a compendium of scRNA-seq datasets. The values displayed represent the average normalized expression across all scRNA-seq datasets.
  • Users have the option to download:
    • The normalized expression matrix used to generate the heatmap.
    • A list of genes associated with a cell state of interest.
  • The genes associated with a cell state of interest may be copied to the clipboard.


Explore Genes Across Cell States

  • Explore genes of interest and their expression across all cell states.
  • Select a gene and a histology to display the expression of that gene in all cell states within the selected histology.
  • An expression value of zero indicates that the gene was undetected in that cell state.

Explore Cell State Survival Associations

  • This heatmap displays the association between cell state abundance and survival.
  • The associations are available as z-scores or directional -log10 P-values. Positive -log10 P-values indicate association with shorter survival, while negative values with longer survival.
  • Explore the cell state expression signature for each association by clicking on the heatmap cells.

Explore Ecotype Survival Associations

  • This heatmap displays the association between lymphoma ecotypes abundance and survival in the discovery dataset.
  • The associations are available as z-scores or directional -log10 P-values. Positive -log10 P-values indicate association with shorter survival, while negative values with longer survival.
  • Explore the cell state expression signature for each association by clicking on the heatmap cells.

Explore Ecotype-Specific Ligand-Receptor Interactions

  • Using this feature, Ecotype-specific ligand-receptor interactions between different cell types can be displayed.
  • Select an Ecotype of interest, then select a celltype whose ligands will be displayed, and a celltype whose receptors will be displayed.
  • Use the button to download a csv file of the currently displayed data.
  • See the downloads tab to download all ligand receptor interaction data for all Ecotypes.

Explore Cell State Networks For Each Lymphoma Ecotype

Analyze Bulk Expression Data

Recover lymphoma cell states and ecotypes in bulk tumor expression data. The maximum size of the input file is limited to 200MB. For de novo discovery of cell states and ecotypes, and unrestricted input matrix size, please use the source code available on EcoTyper GitHub page and in the Downloads section.

See the analysis tutorial for information on how to format input data. See the downloads page to download an example bulk dataset for running EcoTyper.

Analyze Single Cell RNA-seq Data

Assign single cells to an atlas of lymphoma cell states using reference-guided annotation. The maximum size of the input file is limited to 200MB. For full functionality, including ecotype recovery and unrestricted input matrix size, please use the source code available on EcoTyper GitHub page and in the Downloads section.

See the analysis tutorial for information on how to format input data. See the downloads page to download an example single cell dataset for running EcoTyper.

Data Format

Gene expression table:

EcoTyper takes as input a bulk gene expression table from RNA-seq or microarray data with the following formatting requirements:

  • The file must be tab delimited.
  • The dimensions of the matrix must be genes (rows) and samples (columns).
  • The first row of the matrix must specify sample names.
  • The first column of the matrix must specify gene names.
  • Gene names should be standard HUGO symbols.
  • Gene names should be unique.
  • Data values can be TPM, FPKM, log2(TPM), or log2(FPKM).

Input data recommendations:

  • We recommend that EcoTyper be run on > 25 samples
  • The supported cancer type is diffuse large B cell lymphoma (DLBCL).

Example Expression Table:

Outputs

Output page:

  • One tab for each cell type analyzed is generated, in the top part of the output page. Each tab contains two heatmaps, analogous to Figures 2A and 2B in the lymphoma EcoTyper paper. A heatmap depicting the cell type specific expression of cell state signature genes across lymphoma samples from the discovery cohort (Schmitz et al.) is displayed as a reference on the left. A heatmap depicting the expression of cell state signature genes in the user input data is displayed on the right: