Skip to main content
. 2018 Oct 30;18(12):141. doi: 10.1007/s11892-018-1111-4

Table 2.

Strategies to establish pathogenicity of a variant in diabetes genes

Approach Parameter Description
In silico data Genome database searches Search of genome/exome databases e.g. GnomAD to establish mean allele frequency and also assess presence of other variants at affected nucleotide position
Sequence variant databases For example, dbSNP, Exome Variant Server or 1000 genomes NGRL
Mutation database searches Human Mutation Genetic Database search, imports published data on genetic mutations. Limitations, as mutations published may not necessarily prove to be pathogenic
Amino acid change Examines the effect of the amino acid substitution on charge and polarity. A significant change in polarity or charge from the amino acid substitution might be more likely to impair protein function
Species conservation Conservation can be scored using tools such as ConSurf, through which multiple sequence alignments can be undertaken to compare orthologs across species. Essential sites for protein function are likely to be invariant across species (highly conserved)
Software prediction Software prediction models: SIFT, PolyPhen2 and AlignGVGD
Grantham Distance assessed the physico-chemical difference between amino acid properties
Cryptic splice site Online software that can predict whether the variant creates a cryptic splice site or alters an existing one
In vitro data Cellular studies For HNF1A, GCK and KCNJ11, published studies exist that demonstrate effects of variants on protein function. In turn, these can be fed back to affected individuals to establish best treatment options or management approaches. Lack of accessibility limits widespread use
Clinical data Biomarkers/systemic features Examining for features associated with the genetic mutation may help decipher whether a variant is disease-causing or benign.
For example, high-sensitivity C-reactive protein levels are known to be lower in people with HNF1A MODY, so demonstrating undetectable levels in a person with a variant of unknown significance may be a helpful indicator to prove functional effects
Low magnesium levels or evidence of pancreatic exocrine failure in people with HNF1B variants, is another example
A history of neonatal hypoglycaemia and/or fetal macrosomia in HNF4A variants might be helpful
In GCK variants, demonstrating only fasting hyperglycaemia or a history of fetal macrosomia may assist
Treatment response Demonstrating sensitivity to sulphonylurea therapy is a compelling piece of evidence favouring pathogenicity in HNF1A and HNF4A variants
Other monogenic genes are not so easily readily identified in this way
Co-segregation studies Proving the variant segregates with diabetes in a family is compelling evidence in favour of pathogenicity. To achieve this robustly, in an affected kindred, people with diabetes should have the variant and those without the variant an absence of diabetes. Conclusive proof comes from co-segregation in a different kindred to the proband