Introduction
Evaluating the potential functional impact of variants in Mendelian
disease genes is a key component in the interpretation of their clinical
significance. Disease gene databases chiefly contain nonsense,
frameshift indels, and missense variants, in addition to variants that
impact donor and acceptor splice site motifs. In particular, synonymous
variants are often dismissed from variant curation and test reporting
under the assumption that they are “silent” variants. However, these
variants can still impact transcription, mRNA processing and translation
(Sauna & Kimchi-Sarfaty, 2011). Further, intronic variants outside of
the donor and acceptor splice site motifs are mostly disregarded in
clinical testing and/or reporting due to the low sensitivity and
specificity of currently available methods to predict their impact on
mRNA splicing. This negative bias in recording of synonymous and
intronic variants has implications for their inclusion in data analyses
and functional studies in research settings.
Current variant interpretation approaches also generally ignore the fact
that all types of exonic and intronic variants can potentially
affect mRNA splicing (we will term these types of variants as being
“spliceogenic”). Exonic variants initially annotated as synonymous,
missense, nonsense or frameshift based on predicted codon usage can
destroy, enhance or create motifs recognized by the mRNA splicing
machinery (see below). Intronic variants outside the native splice sites
can destroy branchpoint (BP) motifs, or create or enhance the use of
cryptic sites. To improve assessment of variant pathogenicity and
clinical decision-making, it is important to expand variant curation and
reporting to include reliable bioinformatic prediction of
spliceogenicity for variants located outside the donor and acceptor
splice site motifs.