Page 94 - Read Online

Page 4 of 12 Papadodima et al. J Transl Genet Genom 2019;3:7. I https://doi.org/10.20517/jtgg.2018.33

Figure 1. The number of publications per year on Pubmed (until 4th December 2018) using terms “melanoma” and “cutaneous
melanoma” (upper), major landmarks concerning the study of melanoma (lower). NGS: next generation sequencing

use different approaches to extract the desired list of variants, meeting certain criteria. Among the strategies
utilized are heuristic approaches combined with statistical tests, analysis and evaluation of a joint genotype
likelihood, allele frequency or haplotype-based analyses, or exploitation of machine learning methods for
variant classification. Apart from these, there are specialized tools that offer single-sample somatic mutation
calling (lack of normal samples), through association with databases like COSMIC [29,30] and application
of machine learning and statistical algorithms. Table 1 lists most somatic mutation callers based on their
aforementioned strategic approaches.

As a latter step, after obtaining a list of somatic mutations, it is important to distinguish the driver mutations
[79]
which actively contribute to carcinogenesis . This can be accomplished through mutation frequency
analysis, functional impact investigation or machine learning algorithms based on known sets of driver/
passenger genes. Another approach followed is enrichment analysis on known pathways or networks. Table 2
summarizes several tools which focus on driver mutation identification, classified by the strategic approach
used. It is important to mention that distinction of driver/passenger genes faces many challenges mostly
due to lack of annotation, additive effects of passenger mutations or a possible change in roles during cancer
[28]
progression and the development of tumour heterogeneity . In a recent publication, our group presented
a methodology combining functional impact analysis with pathway enrichment, to deal with a limited
dataset, in order to distinguish important genes and possible drivers exploiting exome sequencing data from
[80]
melanoma patients in Greece .

GENES BEARING CAUSATIVE SOMATIC MUTATIONS IN MELANOMA
One of the most well-established pathways commonly affected in melanoma is the mitogen activating
protein kinase (MAPK) signaling cascade, governing cell growth and survival. BRAF, NRAS and NF1 are the