Leveraging recent advances in human genome annotations to identify structural variants associated with cardiometabolic phenotypes in diverse populati

Cardiometabolic diseases (CMDs) collectively remain the leading causes of death worldwide. CMD are a group of common, interrelated conditions including heart disease, stroke, and metabolic disorders like diabetes and obesity. These

diseases are often linked by shared risk factors and underlying mechanisms, such as inflammation, insulin resistance, and dyslipidaemia. While substantial efforts, over the past 15-years, in genetic studies have led to a deeper knowledge on the pathophysiological processes in CMDs and in the identification of novel treatment targets (e.g. through lipid modification), our understanding on the genetic factors underlying CMDs risk still remains incomplete and requires further research.

Particularly, while most studies have focused on the association between disease risk and genetic variants at a single nucleotide, the influence of structural variants (SVs) on cardiometabolic risk is largely unexplored. SV are defined as genomic variations exceeding 50 base pairs, including deletion of sequence, duplication, inversion, translocation, insertion, or tandem repeats-a segment of two or more bases that are repeated numerous times

1. Structural variants can significantly impact health and they've been associated to several diseases including cardiometabolic diseases, autism, schizophrenia, Huntington's disease, and many others

2. SVs are typically detected by mapping short-read sequences from whole genome sequencing to the reference human sequence

3. However, the current human reference is a single mosaic of genetic data assembled from the sequence of more than 20 individuals, containing errors and several gaps in genetic regions difficult to assemble, particularly those that are repetitive. Moreover, a single reference genome is not sufficient to precisely map the diversity often observed between segments of genomes from different individuals

4. Therefore, until recently, over 70 percent of SVs in the human genomes remained unidentified. Very recent efforts from different consortia including the Human Pangenome Reference (HPRC), the Telomere-to-Telomere, and the Chinese Pangenome have generated using long-read sequencing technologies the complete phased, diploid assemblies of several genetically diverse individuals5,6.

These new pangenome reference datasets have started to shed new insights on previously unexplored and inaccessible regions of the human genome. Results from the pangenome reference has already indicated that of the newly identified 119 million nucleotide bases about 90 millions of these derive from structural variation

5. The availability of the Pangenome reference assemblies for short-read sequencing studies has been shown to improve SV calling up to 104% (approximately 13,000 SVs per genome). As such, the ability to accurately determine SVs with these newer reference assemblies, using existing short-read sequencing studies will enable a systematic evaluation of SVs with health and diseases risks, providing novel insights on disease processes.

The primary aim of the study is to leverage recent advances in human genome assemblies to identify SVs associated with cardiometabolic phenotypes and outcomes in existing sequencing studies of diverse population groups

In this project the student will apply existing pipelines and develop new approaches to use pan-genome graph-based assemblies to genotype SVs in human datasets of different ethnicity and:

- Take advantage of whole-genome sequencing data in >800 monozygous twins to assess the concordance of SV distribution in these identical twins, investigate their mutation rate and identify regions and features of the genome more prone to accumulate these mutations

- Use already-generated multi-omics molecular data, clinical and biochemical traits for a deep investigation of the impact of SVs on cardiometabolic diseases