Next-generation sequencing (NGS) at the Children's Hospital Zagreb - what have we learned so far?

Abstract

Next-generation sequencing (NGS) technology has enabled the sequencing of hundreds of genes, and even the entire genome, in a single step, which has initiated a real revolution in the diagnosis of disorders with a hereditary basis. In addition to diagnostics, NGS technology has also enabled significant therapeutic advances, as can be seen in examples of hereditary disorders for which there is targeted treatment (e.g. epilepsy, metabolic disorders, malignant diseases).
The development of NGS technology has also been greatly contributed to by progress in the development of bioinformatics techniques for genome analysis and computer tools that enable the processing of data obtained through sequencing and its analysis and interpretation for clinical and research purposes.
In everyday practice, NGS diagnostics takes place through several phases; from initial sample collection, DNA isolation, creation of a DNA library to the sequencing itself, which produces the so-called raw data that needs to be analyzed. NGS technology is constantly being improved, starting from the automation of certain parts of the protocol and the development of quality control tools to computer tools for variant analysis that are continuously updated with data on new genes, the content of genomic and literature databases.
 Aware of all the benefits of this technology for the health care of patients, we implemented NGS diagnostics in our Clinic more than 7 years ago, initially by introducing clinical exome sequencing (CES) and more recently by implementing whole exome sequencing (WES).
 Although it is a technology that has significantly improved everyday work, for proper application it is important to know its limitations. Namely, it is known that this technology has limited efficiency in the detection of structural variants and sequencing of homologous genes, genes with pseudogenes, GC-rich regions and repetitive regions.
 Using the example of three patients, we will present some of the limitations we encountered in our work and emphasize the importance of knowing additional tools in the interpretation of variants, phenotyping, knowing other diagnostic methods of genome analysis and knowing the mechanisms of disease development at the genomic level.
Example 1. A 17-year-old patient comes for treatment of primary amenorrhea, uterine agenesis and vaginal anomalies. The karyogram performed was 46,XY. The clinical diagnosis of androgen insensitivity syndrome (AIS) was made, which is caused by variants in the AR gene, which encodes the androgen receptor. After analysis using the CES method, the computer tool for tertiary data analysis did not detect pathological changes in the AR gene, but an inspection of the IGV viewer (Integrative Genomic Viewer) revealed the absence of reads in exons 2-8/8 in the AR gene. A chromosomal microarray confirmed a deletion on the X chromosome in the 96 kb region of q12, which affects the aforementioned exons in the AR gene.
Example 2. A 2-year-old female patient presents for treatment of iris coloboma, tetralogy of Fallot, dysmorphism, and developmental delay. CHARGE syndrome is suspected. CES analysis performed according to HPO (Human Phenotype Ontology) codes did not detect the causative pathogenic variant. Subsequent WES analysis revealed a heterozygous likely pathogenic variant (PVS1, PM2) in the CHD7 gene (NM_017780.4): c.6718del, p. Asp2240MetfsTer37. When viewed in the IGV viewer, the variant is visible in both diagnostic platforms (CES and WES), but during the CES analysis, the computer tool rejected it due to too few reads.
Example 3. A 3-year-old female patient presents for treatment of microcytic anemia and macrocrania. The pathogenic variant was detected by the CES method; deletion of 5 nucleotides at position c.95+2_95+6 affecting the splice site in the HBA1 gene and/or the HBA2 gene (PVS1, PP5, PM2). The HBA1 and HBA2 genes are paralogous genes with pronounced homology, which is why it is not possible to define in which of these two genes the specified variant is found with this sequencing method (eng. short read) and whether it is a homozygous or heterozygous variant. Pathogenic variants of both genes are associated with hemoglobinopathies - alpha thalassemia. Long read sequencing technology would enable the definition of zygosity and the gene on which the specified variant is located.