TRAJECTORY AND NETWORK ANALYSES UTILIZING TRANSCRIPTOME DYNAMICS REVEALS HIGHER ORDER COORDINATION EFFECTS IN THE PREIMPLANTATION HUMAN EMBRYO

Abstract

The early human embryo experiences significant changes in trajectories during preimplantation development. This dynamic nature of the embryo is driven by a highly coordinated transcriptome and is connected to the embryonic genome activation (EGA) process. DOHaD (Developmental Origins of Health and Disease) markers can be found within the group of genes which influence the coordination of the transcriptome at this stage. A novel batch correction technique for integration of RNA-velocity compatible data was developed to combine publicly available human preimplantation embryo single-cell transcriptomics data to allow for the construction of a comprehensive cell atlas of human embryo development from fertilization to day 14. Combined trajectory analysis which includes velocity, pseudotime, and trajectory inference was used to characterize the development patterns of the preimplantation human embryo. The transcriptomic modules behind the observed developmental patterns were identified and characterized using a multifaceted approach to network analysis, including the use of hypernetworks to capture complex gene interactions. Finally, findings from previous studies regarding catch-up growth in small for gestational age (SGA) children were brought into the preimplantation period to identify potential risk markers through causal analysis using transfer entropy. Trajectory signatures of vital developmental events were identified and found to line up with currently known development timings including evidence for EGA wave at the 8-cell stage as well as blastocyst differentiation at day 5. Network analysis revealed a disparity in information content between higher- and first-order transcriptomic networks throughout the data and that the differences are more pronounced around differentiation events, suggesting higher level coordination behind them. Hyperconnected genes at morula formation were found to be highly correlated with vital functions enabling the process. 4 genes were identified as potential SGA risk markers and were found to be understudied in the preimplantation period.

Date of Award	12 Jul 2024
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Adam Stevens (Main Supervisor), Peter Ruane (Co Supervisor), Daniel Brison (Co Supervisor) & Peter Clayton (Co Supervisor)