Ovarian cancer (OC) remains a leading cause of mortality in women due to late diagnosis and frequent relapse (70%). While the mutational landscape of the coding genome is nearing completion, less is known about the characteristics of its noncoding genome. Non-coding regions contain regulatory elements essential for gene expression, and mutations in these areas can disrupt cellular functions critical to OC development and disease progression.
Question:
Do specific noncoding mutations contribute to the likelihood of disease relapse in patients?
Objectives:
1. Identify noncoding variants with impact on mRNA levels
We have developed an integrated framework that combines WGS and RNA-Seq to identify noncoding variants with impact on target genes (within the same TAD domain, or near the variant). Here, we would focus on OC patients and group them based on the occurrence of the variants to identify associated changes in gene expression.
2. Identify noncoding variants from objective 1 that predict OC phenotype and clinical outcomes
For time-dependent outcomes, we will employ Cox proportional hazards models to assess survival rates, disease progression and relapse risk, in relation to the presence of specific noncoding variants. We will use logistic regression to analyse the associations between these variants and other clinical variables, including age at diagnosis and cancer stage.
3. Identify histological features associated with predicted variants from objective 2
For the predicted variants identified in objective 2., we will further analyse histological images to identify features associated with the variants. We have previously developed an AI system that maps omics measurements to histological features. We would apply this system to both the patient cohort with the identified variants and those without, to identify predicted features linked to these variants.
Our study will at the end identify unique noncoding variants with functional effects.