Page 14 - Read Online

Page 452 Aledo-Serrano et al. J Transl Genet Genom 2021;5:443-55 https://dx.doi.org/10.20517/jtgg.2021.40

our DNA sequencing ability, there are still difficulties with GC-rich regions, homology, or pseudogenes.
Additionally, although microarrays (array-CGH) show greatly improved ability to confirm CNVs, their
[71]
limited resolution on pre-defined sequences means that some diagnoses can still be missed . Moreover,
there are technical limitations of NGS that are typical in the bioinformatics algorithms used to align
genomic data to a reference sequence, and annotation and filtering strategies may affect which variants are
[72]
subsequently identified for analysis . Whole genome or trio genome (including parents) sequencing
(WGS) studies are starting to be used, with some publications showing promising results in DEEs without
[73]
diagnosis . However, the major difficulty associated with WGS is the arduous mass of information
provided, which must be properly mined and interpreted through close geneticist-clinical collaboration.
The roles of many genes and complex genome variations are still undetermined. Thus, although WGS can
produce a large volume of data, most of them may be misleading or useless at the moment. Then, the
genetic testing strategy selected for a patient with an undiagnosed DEEs should be analyzed in detail. This
will help us to better understand with which probability are current methodologies in genetic screening
missing disease mechanisms .
[74]
Millions of individuals with rare diseases have undergone targeted genetic assessment, and more than
100,000 patients worldwide have undergone an NGS sequencing technique (especially whole exome
[75]
sequencing approach) during their clinical care . However, access to most of these data is difficult, if not
impossible, because healthcare is a highly regulated environment due to policies on personal data privacy
and potential solutions are difficult and costly. Even in research, taking in account that individuals usually
agree to share their whole data, access is usually a very limited process to ensure compliance. Strategies must
be developed to capitalize on previous as well as future data and fully exploit the powerful collective data in
order to assist the interpretation of genetic variation for patient care . The balance between data sharing
[76]
and maintaining confidentiality will ultimately impact our diagnostic capacity in the future.

Finally, we must not lose sight that current clinical-grade NGS and associated bioinformatics analyses
cannot confirm genetic variation associated with non-Mendelian inheritance, such as tissue (brain)-specific
somatic mosaicism (very frequent in epilepsy and already widely linked to cortical developmental
malformations), epigenetic abnormalities, oligogenic mechanisms, or gene-environment interactions . For
[77]
instance, pathogenic variant mosaicism might not be easily identified depending on the accuracy of the
selected approach, and detection is especially hampered for tissue-specific mosaicism (especially in the
central nervous system) and low-level somatic mutations. Epigenetic alterations, such as methylation
defects, uniparental disomy, and altered expression profiles of the parents of origin at a locus, require
specific testing approaches. In addition, there are currently no clinical-implemented strategies to detect
oligogenic or complex inheritance (even with polygenic risk scores), where the cumulative effect of different
common and rare variants, together with environmental triggers in the context of a complex disease, results
in the clinical picture of a specific patient . The study of DEEs caused by these mechanisms is currently
[78]
limited and will warrant attention and research in the next future.

DECLARATIONS
Authors’ contributions
Made substantial contributions to conception, material articulation, and editing of the review: Aledo-
Serrano A, Sánchez-Alcudia R, Tolerano R, Garcia-Morales I, Beltrán-Corbellini A, del Pino I, Gil-Nagel A
Provided administrative, technical, and material support: Aledo-Serrano A, Sánchez-Alcudia R
Provided material support: Sánchez-Alcudia R, Tolerano R, Garcia-Morales I, Beltrán-Corbellini A,
del Pino I, Gil-Nagel A