Data Specifications

Need to move your cytometry data from Teiko into another clinical trial analysis platform? These are the data specifications you need.

This example is from the data transfer specification from app.teiko.bio demo dataset UMN002.

Definitions

Gating: The process of classifying cell events into biologically descriptive cell populations and cell states based on marker expression. This involves drawing boundaries or “gates” over specific cell populations to classify those cell populations’ respective lineages.

Clustering: The process of using unsupervised machine learning to identify cell types defined by their expression of markers without prior knowledge of cell populations.

FCS: Flow Cytometry Standard, a .fcs file format for cytometry data containing mass cytometry acquisition events and their associated channel values. The files provided will be version 3.1. Specifications for this format can be found here.

GatingML: a GatingML file format used to store a representation of cytometry gates. The file format version to be provided will be version 2.0. Specifications for this format can be found here.

Data Description and Usage

Teiko Bio’s analysis produces one FCS file per sample and one GatingML2.0 file per sample.

  • FCS file per sample. Will provide a directory containing all files per sample.UMN002_fcs_YYYY_MM_DD_HH_MM_SS<sample_name>.fcs
  • GatingML per sample. Will provide a directory containing all files per sample.UMN002_gatingml_YYYY_MM_DD_HH_MM_SS<sample_name>.gatingml

We generate up to five results files per analysis summarizing different outputs of our analysis pipeline.

Each file contains metadata to identify results associated with a particular sample and subject.

Core customers receive three files:

  1. Gated population frequencies by sample (UMN002_gated_population_frequencies_by_sample_YYYY_MM_DD_HH_MM_SS.csv)
  2. Gated population cell state frequencies by sample (UMN002_gated_cell_state_by_sample_YYYY_MM_DD_HH_MM_SS.csv)
  3. Gated population marker expression by sample (UMN002_gated_marker_expression_by_sample_YYYY_MM_DD_HH_MM_SS.csv)

Premier customers receive these three files, and two more:

  1. Clustered population frequencies by sample (UMN002_clustered_population_frequencies_YYYY_MM_DD_HH_MM_SS.csv)
  2. Clustered population marker expression by sample (UMN002_clustered_marker_expression_YYYY_MM_DD_HH_MM_SS.csv)

For example, if a Core customer sends Teiko four cohorts, with nine samples in each cohort (36 samples total), the client will receive 36 FCS files, 36 GatingML files, and three results files.

If a Premier customer sends Teiko four cohorts, that customer will receive the three results files (Gated population frequencies by sample, gated population cell state frequencies by sample, gated population marker expression by sample), and in addition, will receive two more (clustered population frequencies by sample, clustered population marker expression by sample).

File Structure

Each table represents distinct data outputs from our analysis pipeline. These tables capture different aspects of cell populations and marker expressions across both traditionally gated and machine-learning clustered cytometry results.