Ayesha Chowdhury, Shashikanth Marri, Lucy T. Anastasi, Alex Ashenden, Tomas Rozek, Jinghua Feng, Lucas Dejong, Rosalie Kenyon, Dominik C. Kaczorowski, Hung Nguyen, Khoa Lam, Kirsty Stallard, Tracy Merlin, Enzo Ranieri, Sunita De Sousa, Nicholas W. G. Smith, Abhi Kulkarni, Ben Saxon, Drago Bratkovic, Christopher Barnett, Carol Wai-Kwan Siu, Hamish S. Scott, Jovanka King, Karin S. Kassahn
Newborn screening programs are instrumental in the early detection of treatable conditions in the first days of life. By integrating genomic approaches, there is potential to expand the range of conditions included in these programs. As part of a research study, NewbornsInSA, we validated a genomic newborn screening workflow. Analysis of whole-genome sequencing data was restricted to a virtual panel of 613 genes, selected through engagement with local clinical teams. We assessed the workflow's performance using retrospective samples with known variant status. To reduce manual curation time, bioinformatics scripts were developed to auto-classify cases into those with no findings and those requiring manual review. We report on early findings from applying this workflow to a prospectively recruited cohort in which five reportable findings have been made to date. We discuss reporting challenges encountered in genes associated with multiple conditions, with incomplete penetrance, or variants associated with only mild phenotypes.