IE Research Datalab members Nacho Molina and Idris Kouadri have published a new preprint on cell-cycle regulation and pluripotency, highlighting work led by Maulik K. Nariya.

The study brings together single-cell RNA sequencing data from six human and mouse pluripotent systems, spanning different naïve and primed states and culture conditions. From those data, the team estimated a continuous cell-cycle phase for each cell, revealing periodic gene-expression dynamics across systems.

To do this, the authors developed Ciclopes, a biology-informed deep-learning framework that uses an autoencoder to reconstruct periodic gene-expression dynamics and estimate phase-dependent transcription and degradation rates.

One of the central findings is that mouse and human pluripotent systems appear to use distinct regulatory regimes for cell-cycle control. Mouse systems maintain a higher baseline expression of core cell-cycle regulators, while human systems show larger oscillatory amplitude with a lower baseline.

The work also points to a surprising layer of conservation: mRNA degradation timing is more conserved than transcriptional timing. The phase at which genes reach peak degradation correlates across organisms more strongly than the phase of peak transcription.

Finally, the team generated a single-cell RNA-seq time course of human induced pluripotent stem cells differentiating into definitive endoderm, showing that lineage commitment remodels the cell cycle in a coordinated way.

Read the preprint: Comparison of cell-cycle gene expression dynamics and mRNA kinetics across mouse and human pluripotent systems.