Bioinformatics Section

Baylor College Of Medicine, Houston TX

Investigators

Linked publications, trials & patents

Abstract

4. ABSTRACT - Bioinformatics Section (BIS) The Center for Precision Medicine Models (CPMM) Bioinformatics Section will meet the bioinformatics needs of the project, building on the infrastructure established during the first stage. To address the increasing use of whole genome sequencing (WGS), we will extend variant annotation and prioritization to also include non-coding gene regulatory variants and provide informatic methods and tools for multi-omic profiling of variant effects in embryonic and adult mice. To improve CPMM efficiency, we will extend the Submission Tracking System by a collaborative Data Sharing Portal and upgrade the Project Management System and LIMS. We will also extend the CPMM Linked Data Hub (LDH) with a new AI-based literature mining pipeline based on the Retrieval- Augmented Generation (RAG) strategy, co-developed with ClinGen, that will identify mentions of submitted variants in literature and automate extraction of information relevant for variant evaluation and modeling. We will additionally implement a patient omics re-analysis pipeline to help resolve a fraction of pathogenic variants through patient omic data alone, without any animal modeling. While the interpretation of potentially pathogenic non-coding gene regulatory variants is much less established than interpretation of coding variants, efforts are underway in the clinical community to develop standards for evaluating their pathogenicity in clinical contexts. To facilitate such variant assessment and modeling, we will collate relevant epigenomic and gene regulatory information in human and mouse. Several CPMM submissions involve modeling congenital anomalies, early- onset phenotypes likely to cause perinatal lethality in mouse, or studies of in utero therapeutics, highlighting the need for embryo phenotyping capabilities. Therefore, we will evaluate transcriptomics and metabolomics data to characterize embryonic and adult phenotypic characteristics resulting from modeling perturbations. In summary, to increase the volume and quality of variant information produced by CPMM the Bioinformatics Section will implement processes and upgrade infrastructure to improve interactions with users, increase Center throughput, and better coordinate with other centers. Working closely with the Disease Modeling Unit and the Preclinical/Co- Clinical Section, we will develop resources to support variant assessment and modeling for both coding and regulatory variants. Given that almost half of CPMM submissions involve modeling of genetic variants that cause perinatal lethality in mice, we will also implement computational methods to identify multi-omic footprints of pathogenic variants of embryonic and adult mice. 1