Microenvironment-informed inference of transcriptional progression geometry

The BIOCURRENT research introduces a novel causal inference framework designed to reconstruct the unique progression geometry of individual donor transcriptomic data. This framework considers various factors including baseline characteristics, microenvironmental context, and latent pseudotime to evaluate transcriptional state transitions.

One key aspect of BIOCURRENT is its ability to compare compressed or expanded progression intervals across different conditions, a process facilitated by the introduction of the DeltaDelta T estimator. This estimator quantifies the differences in pseudotime intervals across various conditions, allowing researchers to assess potential changes in these intervals when microenvironmental programs are hypothetically modulated.

The framework has been applied to study thymic T-cell developmental lineages as well as immune dysregulation in COVID-19 patients, revealing condition-specific and donor-specific distortions in the progression intervals. By conducting counterfactual simulations, BIOCURRENT links microenvironmental context to alterations in specific intracellular state transition intervals.

The study further localizes these deviations in pseudotime geometry, pinpointing whether changes in transcriptomic programs occur earlier or later along the transcriptomic coordinates. This localization is pivotal in supporting transcriptional stage-aware mechanistic hypotheses and identifying potential intervention checkpoints within complex biological systems.