Table 1.
Summary descriptions of the curation confidence scores linked to the variant combinations present in OLIDA. For each type of evidence, if the information found for a combination does not fulfil the criteria to provide at least a Weak (1) score, an Absent (0) score is assigned. Decision trees (see Supplementary Data, File 1) are used to define the FUNmanual, FUNmeta, FINALmanual and FINALmeta scores. The GENEmanual_harmonized score is defined as the best GENEmanual score assigned for a gene pair among the research articles. More details on how each confidence level is defined per evidence can be found in the Supplementary Data, File 1
| FAMmanual: familial evidence based on the article | ||
|---|---|---|
| Weak (1) | Moderate (2) | Strong (3) |
One of two conditions:
|
One of two conditions:
|
Information of healthy first- and second-degree relatives, showing a perfect segregation of the variants according to the phenotype |
| STATmanual: statistical evidence based on the article | ||
| Weak (1) | Moderate (2) | Strong (3) |
| Implicit evidence that healthy individuals do not carry the oligogenic combination based on control cohorts or public databases. Known control phenotypes, sufficient control size and matched ethnicity | Explicit evidence that healthy individuals do not carry the oligogenic combination based on control cohorts and public databases. Known control phenotypes, sufficient control size, matched ethnicity and (preferably) similar sequencing technology | NA |
| STATknowledge: statistical evidence based on databases and cohorts | ||
| Weak (1) | Moderate (2) | Strong (3) |
| The combination is not found in the 1000 Genomes Project and relevance of all involved variants in ClinVar | NA | NA |
| STATmeta: maximum of STATmanual and STATknowledgea | ||
| Weak (1) | Moderate (2) | Strong (3) |
| The oligogenic combination is not found in the 1000 Genomes Project. Other implicit evidence of its statistical relevance for the phenotype | The variant combination is not found in the 1000 Genomes Project. Additional explicit evidence that healthy individuals do not carry the oligogenic combination based on control cohorts or public databases of matched ethnicity and sufficient control size | NA |
| GENEmanual: gene functional evidence based on the article | ||
| Weak (1) | Moderate (2) | Strong (3) |
| Relevance of involved pathway(s) or expressed tissues on the studied phenotype | One of two conditions: a) Effect of the gene combination on the observed phenotype using a functional experiment with either only a double knock-out or multiple single-gene knockouts b) Direct gene relationship (e.g. common pathway and direct interaction) and relevance for the studied phenotype. | Synergistic or additive effect of the gene combination on the observed phenotype using a functional experiment with single and multiple gene knockouts |
| GENEknowledge: gene functional evidence based on databases | ||
| Weak (1) | Moderate (2) | Strong (3) |
| Relevancy of Reactome or KEGG pathways linked with the genes for the observed phenotype. | One of two conditions: a) Gene combination forms a connected PPI network and the comPPI score of each link is >0.8 b) Common Reactome or KEGG pathways, relevant for the observed phenotype. | NA |
| GENEmeta: maximum of GENEmanual_harmonized and GENEknowledge | ||
| Weak (1) | Moderate (2) | Strong (3) |
| Relevance of the genes on the studied phenotype using pathway or tissue expression information | Direct gene relationship or effect of the gene combination on the observed phenotype without comparing the individual effects of genes | Synergistic or additive effect of the gene combination on the observed phenotype using a functional experiment with single and multiple gene knockouts |
| VARmanual: variant functional evidence based on the article | ||
| Weak (1) | Moderate (2) | Strong (3) |
| One of three conditions: a) All variants are predicted as pathogenic b) Functional experiments for some variants and predicted pathogenic effects for the rest c) Functional experiments using single-variant mutants for the involved variants with a promising but not conclusive effect on the observed phenotype | One of two conditions: a) Effect of the variant combination on the observed phenotype using a functional experiment with either only a double mutant or multiple single mutants b) Clear pathogenic impact of the variant combination on the observed phenotype in an in silico analysis of the joint effect of the variants | Synergistic or additive effect of the variant combination on the observed phenotype using a functional experiment with single and multiple gene mutants |
| VARknowledge: variant functional evidence based on predictors | ||
| Weak (1) | Moderate (2) | Strong (3) |
| Pathogenicity prediction for all variants by at least one predictor among CADD, SIFT, MutationTester and Polyphen | NA | NA |
| VARmeta: maximum of VARmanual and VARknowledge | ||
| Weak (1) | Moderate (2) | Strong (3) |
| Pathogenicity predictions for all involved variants or inconclusive effects of functional experiments | One of two conditions: a) Effect of the oligogenic combination on the observed phenotype using a functional experiment with either a double mutant or multiple single mutants b) Clear pathogenic impact of the oligogenic combination on the observed phenotype in an in silico analysis of the joint effect of the variants | Synergistic or additive effect of the variant combination on the observed phenotype using a functional experiment with single and multiple gene mutants |
| FUNmanual: functional evidence based on GENEmanual and VARmanual FUNmeta: functional evidence based on GENEmeta and VARmeta | ||
| Weak (1) | Moderate (2) | Strong (3) |
| Based on a decision tree, not enough evidence to suggest synergy, but relevance of the involved genes and variants | Based on a decision tree and evidence of a relationship, as well as potential functional synergy for the involved genes and variants, but the joint pathogenic effect on the studied phenotype is still not confirmed or clear | Based on a decision tree, strong evidence of the functional synergy of both involved genes and variants on the studied phenotype |
| FINALmanual: overall evidence based only on Manual scores FINALmeta: overall evidence based on Manual and Knowledge scores | ||
| Weak (1) | Moderate (2) | Strong (3) |
| Based on a decision tree and evidence of the relevance of the variant combination for the observed phenotype but not enough to show that the involved variants are the only culprits for the studied phenotype or that the cause is indeed oligogenic | Based on a decision tree, good genetic and functional evidence of an effect of the oligogenic variant combination on the observed phenotype, but the described information/mechanism is not clear or strong enough to provide proof of oligogenicity | Based on a decision tree and strong evidence of the synergistic/additive effect of the oligogenic variant combination on the observed phenotype genetically and functionally |
a
Exception: if STATknowledge = 0 (because the variant combination is found in the 1000 Genomes Project), then it replaces STATmanual and, thus, STATmeta is also 0.