Modeling the cascade of genome rearrangements in the ciliate oxytricha trifallax

The National Institute of General Medical Sciences (NIGMS) of the Department of Health and Human Services awarded The Trustees of Columbia University in the City of New York $227,652 on June 26, 2026, under the Biomedical Research and Research Training program (CFDA 93.859) to model the cascade of genome rearrangements in the ciliate Oxytricha trifallax.

The recipient will develop quantitative summaries of rearrangement progression using unsupervised learning to uncover how sequence complexity and regulatory features shape timing and pathway dynamics; construct and test competing stochastic and deterministic models to identify principles governing rearrangement order and timing; and identify, classify, and validate naturally occurring rearrangement errors through PCR/QPCR analysis, exploring sequence features that explain error susceptibility. The research leverages a first-of-its-kind, time-resolved, long-read sequencing dataset capturing genome-wide rearrangement intermediates across development and will define generalizable rules for accurate genome reassembly and lay a foundation for understanding evolutionary origins of genome scrambling.

Performance occurs in New York, New York. The award runs through May 31, 2029.

Abstract

Genome rearrangement is a widespread process in development and evolution that includes translocations, inversions, deletions, and duplications. Organisms can benefit from developmental genome rearrangement processes, which enable access to otherwise inaccessible regions of the genetic landscape and allow for combinatorial variation. However, rearrangement can also introduce genetic errors that impair normal gene function and lead to genome instability and diseases such as cancer. The ciliate oxytricha trifallax is an excellent model organism for studying mechanisms and fidelity of programmed DNA rearrangement because of its complex genomic structure. Its transcriptionally active genome is massively reorganized from a scrambled germline template during development, requiring the precise deletion, inversion, and translocation of tens of thousands of sequence segments. This project leverages a first-of-its-kind, time-resolved, long-read sequencing dataset that captures genome-wide rearrangement intermediates across development. In aim 1, i will develop quantitative summaries of rearrangement progression and use unsupervised learning to uncover how sequence complexity and regulatory features shape timing and pathway dynamics. In aim 2, i will construct and test competing stochastic and deterministic models to identify the principles governing rearrangement order and timing. In aim 3, i will identify and classify naturally occurring rearrangement errors, validate key intermediates with PCR/qpcr, and explore sequence features that may explain error susceptibility. This research will define generalizable rules for accurate genome reassembly, providing a model for understanding how genomes self-organize with high fidelity. IT also lays a foundation for future comparative work on the evolutionary origins of genome scrambling and may inform broader principles of genome regulation, editing, and stability in both health and disease.