behavior guided transcriptomics

In this project we aimed to link behavioral characteristics of engineered T cells with a particular transcriptome.

Type of data: Two data types were used and inferred together. (1) Imaging time series data from engineered T cells where two cells types CD4+ and CD8+ T cells were identified by different colors. (2) Single cell sequencing on engineered T cells harvested from a co-culture with patient derived tumoroids. T cells were harvested at different conditions in order to discriminate cell environment and behavior.

Background: In our assay we investigated how T cells where behaving during the course of tumor attack. Additionally we could isolate these cells from the assay and acquire gene expression information about single cells in this assay.

Goal: We aimed to predict a behavioral signature of single T cells for which gene expression data was available.

Approach: Using the fundamental principle of transitivity of probabilistic distribution, I build a probability map distribution for different behavioral signatures over pseudotime arranged T cells.

Main result: From the whole pool of heterogenous genes expression profiles we identified the unique transcriptome of T cells during the course of a tumor attack.