Files and links (20)
Data (supplemental)Datasets S1 to S8 and Movies S1 to S3 Chapter 4Open Access (License Unspecified), Open Access
Dissertation
CYCLOPS 2.0: improving and applying circadian phase reconstruction in confounded datasets
Doctor of Philosophy (Ph.D.), Drexel University
May 2025
DOI:
https://doi.org/10.17918/00011113
Abstract
Circadian rhythms regulate the timing of thousands of genes across human tissues, aligning physiology with the 24-hour day. However, most transcriptomic datasets lack time-of-day annotations, limiting our ability to study these rhythms and their disruption in disease. We improve CYCLOPS, an unsupervised algorithm for inferring internal circadian time from gene expression data, by introducing CYCLOPS 2.0--a covariate-aware model that adjusts for batch effects and non-circadian confounders simultaneously. Benchmarking confirms its improved accuracy in recovering latent circadian structure under noisy conditions. We apply CYCLOPS 2.0 to breast cancer and GTEx datasets to reconstruct circadian transcriptional order. In breast tumors, especially luminal A, we observe persistent but reprogrammed rhythms, with enhanced cycling in EMT and immune pathways. A novel metric, CYCLOPS magnitude (CMag), quantifies global rhythm strength and correlates with reduced five-year survival. Functional assays confirm that circadian disruption reduces invasiveness, linking molecular rhythms to metastatic behavior. We identify circadian eQTLs (cQTLs)--variants that affect rhythmic parameters of gene expression--in adipose, muscle, and skin. Many cQTLs are not detected by traditional methods and colocalize with GWAS loci and circadian transcription factor motifs. These findings reveal a temporally dynamic layer of gene regulation with clinical and functional relevance.
Metrics
264 File views/ downloads
86 Record Views
Details
- Title
- CYCLOPS 2.0
- Creators
- Jan Alexander Hammarlund
- Contributors
- Ron C. Anafi (Advisor) - University of PennsylvaniaAndres Kriete (Advisor) - Drexel University, School of Biomedical Engineering, Science, and Health Systems
- Awarding Institution
- Drexel University
- Degree Awarded
- Doctor of Philosophy (Ph.D.)
- Publisher
- Drexel University; Philadelphia, Pennsylvania
- Number of pages
- xiii, 207 pages
- Resource Type
- Dissertation
- Language
- English
- Academic Unit
- School of Biomedical Engineering, Science, and Health Systems (1997-2026); Drexel University
- Other Identifier
- 991022064341804721