Suzanne de Bruijn

291 General discussion and perspectives (first generation sequencing) or WES do allow the detection of copy number changes, these methods are relatively insensitive as breakpoints of SVs cannot be readily determined and inversions or translocation events cannot be detected. As described in chapter 4 , where RP17-associated SVs were characterized, breakpoint resolution has appeared to be crucial for variant interpretation. SVs can lead to disruption of the 3D chromosome landscape; new regulatory domains can be formed (neo-topologically associating domains (TADs)), regulatory domains can fuse (TAD-fusion) or be deleted (TAD-deletion). Only when the breakpoints of an SV are precisely determined, the effects on the 3D chromosome landscape, and consequently enhancer-promoter interactions, can be correctly interpreted. Therefore, the arrival of WGS has been essential to resolve the genetic mysteries underlying RP17. The findings in chapter 4 indicate that SVs can be an important and unrecognized cause of inherited RD and HL. An important contribution of SVs to the genetic spectrum of both RD and HL has been recently suggested in literature as well. 45,46 Based on this, it can be speculated that by expanding the application of WGS to clinical practices, a significant number of pathogenic SVs will be reported in the next years. Third generation sequencing Although SV-detection has been significantly improved by the implementation of WGS, there is still room for further improvement. As extensively described in chapter 1.2 , short-read WGS approaches are inefficient in the detection of complex SVs. Long-read sequencing techniques (single-molecule real-time (SMRT) (Pacific Biosciences) and nanopore sequencing (Oxford Nanopore)) hold the promise to allow detection of all SVs. Additionally, these techniques have a superior performance in the analysis of repeat- rich and homologous regions, which is an important drawback of short-readWGS. Most likely, the implementation of these techniques is necessary to determine the complete genetic variation landscape in an individual. On the other hand, SMRT and nanopore sequencing still include a relatively high false-positive rate for the detection of single nucleotide variants and small insertions and deletions (≤15 bp) and therefore are still unfit to fully replace short-read WGS approaches . Nevertheless, the implementation of HiFi reads, also called circular consensus reads, in SMRT sequencing promises to improve the accuracy up to 99% (reviewed in chapter 1.2 ). The future of variant detection With the implementation of innovative sequencing techniques, and consequently, the generation of larger sequencing datasets, an increasing need and a rapid development of additional computational tools can be anticipated. Artificial intelligence software,