NGS Cloud is a clinical decision support system that automatically presents a list of candidate variants in the light of submitted phenotypic data, with minimal or no need for filtering to reach a clinical diagnosis. The most fundamental component of this system is the GENIUS algorithm, which performs gene-genotype-phenotype matching and variant prioritisation.

GENIUS (GENe Interactions Under Scope)

Approximately 70,000 variants in whole-exome sequencing (WES) and 5,000,000 variants are called in whole-genome sequencing (WGS) (1). These variants are then filtered and excluded according to several different parameters, such as

• the observed allele frequency in population-based studies, 
• the prevalence of the disorder, 
• the phenotypic relationship(s) of the gene, 
• the overlapping degree of the phenotype with the samples’ phenotype, 
• the known inheritance pattern and the compatibility of the genotype, 
• the functional evidence in the literature, 
• the familial segregation status, 
• whether it has been detected in similarly affected individuals, 
• computational & predictive effects on the protein.

Here, the GENIUS algorithm has been developed to automatically apply all known relevant criteria and enables you to reach the disease-causing variant faster and easier.

Inputs of the GENIUS algorithm:

1. Human Phenotype Ontology (HPO) terms describing human diseases with a standard vocabulary
2. HOPE (Harmonization of Ontologies by PairEnd) terms enriched with several different ontologies.
3. Current age and age of onset
4. Sex

In a recent validation study, GENIUS has presented six variants during a WES analysis on average (Figure 1) (2).

Figure 1. Automatically listed candidate variant numbers

The GENIUS algorithm listed clinically relevant variants in 94.4% of the cases. These variants were either the first or among the top three in 44.1% and 73.5% of the samples, respectively2. The GENIUS algorithm enables a straightforward analysis through tens of thousands, millions of genetic variants and identify genetic changes that cause disease, usually one or two. Most importantly, it becomes more accurate by reducing human-induced errors.

References:

1. Alfares, A., Alsubaie, L., Aloraini, T., Alaskar, A., Althagafi, A., Alahmad, A., Rashid, M., Alswaid, A., Alothaim, A., Eyaid, W., Ababneh, F., Albalwi, M., Alotaibi, R., Almutairi, M., Altharawi, N., Alsamer, A., Abdelhakim, M., Kafkas, S., Mineta, K., Cheung, N., … Alfadhel, M. (2020). What is the right sequencing approach? Solo VS extended family analysis in consanguineous populations. BMC medical genomics, 13(1), 103. https://doi.org/10.1186/s12920-020-00743-8
2. http://www.medicaljournal.gazi.edu.tr/index.php/GMJ/article/view/2914