Chen Lab Research Projects
The Multi-Omics for Mothers and Infants Consortium (MOMI) Scale-up: Genomics
Pregnancy phenotypes such as gestational duration, fetal growth, and maternal physiology collectively shape newborn and maternal health. The Bill & Melinda Gates Foundation has funded the MOMI consortium to optimize health for mothers and infants in low-and-middle-income countries using a multi-omics approach. In this study, we will collaborate with Dr. Ge Zhang and all MOMI international collaborators to (1) generate genome-wide genotyping data of 25,400 maternal and fetal DNA samples using low-pass whole genome sequencing, and (2) conduct comprehensive genomic analyses of multiple pregnancy phenotypes in diverse populations.
Pediatric Cardiac Genomics Consortium Genomics Data Center
The Pediatric Cardiac Genomics Consortium (PCGC) is a group of clinical research teams formed by Boston Children’s Hospital, Columbia University, University of Utah, and Yale University and supported by NHLBI, collaborating to identify genetic causes of human congenital heart disease and to relate genetic variants present in the congenital heart disease to clinical outcomes. As part of the Administrative Coordinating Center led by Dr. Michael Wagner, our genomic data group’s responsibility includes managing, processing, and analyzing WES, genotyping array, and WGS data derived from 35,000 individuals including patients and their family members in the PCGC cohort.
Noonan Syndrome Detection Based on Deep Learning of Electronic Health Records
Noonan syndrome is the most common syndromic cause of congenital heart defect second only to Down syndrome. It can affect 1 in every 1000 to 2500 live births in the United States. Recent studies have raised the concern that there could be substantial number of undiagnosed Noonan syndrome in the population. In this CpG funded project, we are collaborating with Dr. Nicole Weaver to (1) develop a deep learning method to detect Noonan syndrome based on electronic health records, and (2) apply the method to identify undiagnosed patients of Noonan syndrome in our hospital. This study will not only demonstrate the validity of our method, but also exemplify the potential of repurposing electronic health records for rare genetic disease screening.
Identifying Patients of Aromatic L-amino Acid Decarboxylase (AADC) Deficiency Based on Electronic Health Records
Aromatic L-amino acid decarboxylase (AADC) deficiency is an ultra-rare genetic disorder caused by mutation in DDC gene. Patients with AADC deficiency have deficient synthesis of dopamine and serotonin, resulting in symptoms like weak muscle tone, developmental delay, and oculogyric crisis. This disease is so rare that less than 200 patients have been reported in the literature worldwide. Although the true incidence is unknown, it’s suspected that there are patients with this disease who are not diagnosed due to lack of recognition. In this pilot study, we are collaborating with Discover Together Biobank to identify patients of AADC deficiency based on electronic health records using a computational method named GDDP that we developed before.
Genetic Diagnoses in a Cohort of Individuals with Valvar Pulmonary Stenosis
Valvar pulmonary stenosis (vPS) is a heart valve disorder which causes a reduction of blood flow to the lungs. It’s estimated to represent 10% of congenital heart defects. To gain more insights about Mendelian and complex genetic etiologies of vPS, and funded by NIH Gabriella Miller Kids First program, we will collaborate with Dr. Nicole Weaver to conduct whole genome sequencing on a cohort of 900 individuals including patients with vPS and their parents. We will perform comprehensive genomic and bioinformatics analysis to identify pathogenic variants in the probands and novel candidate genes that are associated with vPS, including phenotype-guided pathogenic variant prioritization using GDDP and disease gene prioritization using ToppGene that we developed before.
