Biomedical Engineering Reference
In-Depth Information
Inheritance in Man (OMIM;
http://www.ncbi.nlm.nih.gov/omim
), Kyoto
Encyclopedia of Genes and Genomes (
http://www.genome.jp/kegg/
), and Ingenuity
System Pathway Analysis (
http://www.ingenuity.com/products/ipa
). In addition to
pathogenicity predictions, fi ltering by using phenotype databases in consideration
of the patient's clinical features (OMIM) and patterns of inheritance among family
members are usually required to identify candidates of deleterious mutations (1000
Genomes Project Consortium et al.
2010
). However, designation of pathogenicity to
variants is still a challenging task.
2.8
Common Sources of Errors in NGS Data
The presence of systematic false positive and negative variation troubles NGS data.
Specifi cally, incorrect genome mapping, equipment sequencing errors, or sequenc-
ing calls of DNA fragments near the ends are responsible for false positives variants
(Lin et al.
2012
). For example, pyrosequencing NGS systems have systematic errors
at 5-6 nucleotide homopolymer stretches (Margulies et al.
2005
). To improve the
accuracy, such error can be removed from the fi nal list of variants. Furthermore,
software alignment artifacts may be reduced by cross software alignment compari-
sons. In contrast, depth of coverage, poor capture effi ciency, and diffi culty in unam-
biguously aligning repetitive regions usually cause false negative variants (Nothnagel
et al.
2011
). This NGS error can be reduced by raising the threshold of the sequenc-
ing coverage to reach clinical standard accuracy such as that of Sanger sequencing
(Koboldt et al.
2010
). Additionally, increasing coverage reduces the error rate, spe-
cifi cally, for heterozygous variants. Moreover, false positive variants can be further
reduced by using longer reads which consequently decreases mapping ambiguity.
When longer reads are used, it is advisable to trim the end nucleotides since they
have higher error rates (Ledergerber and Dessimoz
2011
). Paired-end sequencing
permits alignment software to fi nd the correct target region and to reduce the inter-
ference from pseudogene regions (Nielsen et al.
2011
). However, for NGS clinical
sequencing panels Sanger sequencing is usually required before the fi nal report can
be issued.
2.9
Clinical Applications of NGS
Despite its errors, NGS has been employed by the research community to identify
causal genes and clinical laboratories have started to apply these technologies for
the diagnosis of Mendelian disorders (detailed in Chaps. 4-8). Most Mendelian
disorders are caused by exonic or splice-site mutations that alter the amino acid
sequence of the affected gene. The number of known mutations in human genes
underlying or associated with inherited disease exceeds 110,000 in more than 3,700
Search WWH ::
Custom Search