A sequencing approach to refine the PEAR1 locus in platelet aggregation

Bioinformatics Internship Presentation

Prabhakaran Gangadharan (Mentor: Dr. Rasika Mathias, Division of Allergy and Clinical Immunology, Johns Hopkins University School of Medicine)

August 26th, 3:00pm-3:20pm Room 1300, Harris Building.

Platelet aggregation is a critical step in the initiation of thrombosis of the arterial system, which results in acute thrombosis-mediated ischemic syndromes such as myocardial infarction, stroke, and peripheral arterial occlusions. Genetic variation is thought to contribute to variability in platelet function; however the specific variants and mechanism that contribute to alter platelet function are poorly defined. The GeneSTAR group (Genetic Study of Aspirin Responsiveness) has previously identified an intronic variant in the PEAR1 gene using a genomewide association study of ~1 million single nucleotide polymorphisms (SNPs). The present study was designed to identify the specific variants within the region of the genomewide study that accounts for this association relying on sequence data in this region on a subset of the GeneSTAR subjects.

Whole genome sequencing (WGS) data from GeneSTAR research program was used for this study. A 25KB region upstream and downstream of PEAR1 gene from chr1 (genetic coordinates chromosome 1: 156893704 to 156916434, NCBI) for 95 African American and 109 European American individuals was subset from the full WGS data using VCFtools. Annotation of all sequence variants was performed using several alternative annotation tools (SeattleSeq, PloyPhen, SIFT). Variants in the PEAR1 gene were classified as rare and common based on minor allele frequency (MAF < 5% is rare, MAF >= 5% is common) and these classified variants were analyzed for association with clinical phenotype PRPAE22AS (platelet-rich-plasma in response to epinephrine adjusted for age and sex) using PLINK/SEQ. Fisher’s exact test was performed for rare variants and logistic regression for common variants.

As a result of sequencing analysis of the PEAR1 gene, a common variant (rs2644592) was identified in intron 5 which shows association with the clinical phenotype in both African American and European American. But the variant rs2644592 is not the same in the earlier genomewide association study, therefore the sequence data may have helped uncover a variant that has a stronger signal than the original genomewide identified locus that is the same in both African American and European American.