Direct Marker Tests 

Most of Wisdom Panel’s™ disorder and trait tests are direct-marker tests, meaning the presence or absence of the causal variant in question is directly assayed. This is considered the gold standard for simple Mendelian disorders and traits, as it provides the highest level of accuracy. As Wisdom Panel™ has high quality assurance standards, direct marker tests are consistent and carry a high level of confidence, typically above 99% accuracy.  

Linked Marker Tests 

Linked-marker tests, also called indirect marker tests, are those in which a marker very close to the marker in question is tested. When very close together on the chromosome, the markers are said to be “linked” and are usually inherited together, giving a reasonable indication of whether the disorder or trait variant was inherited. When Wisdom offers a linked-marker test, it is most commonly when a direct test is not possible because the mutation location or type is not compatible with our testing technology, or the causative variant is not yet known, but there is urgent need for a preliminary test. 

Wisdom carefully considers linked-marker test performance before offering them in reports, to ensure the linked-marker test is highly predictive, and clearly marks any linked-marker tests in reports, so there is never a question of test type. Linked-marker tests vary in reliability depending on testing laboratory, testing technology, marker, marker location, and likelihood of linked inheritance, so it is important that the laboratory in question has a commitment to high quality standards. Although Wisdom carefully evaluates their linked-marker tests for excellent performance, with any linked-marker test the possibility of an incorrect call is more likely than with direct tests, and so they should be interpreted accordingly. 

Predictive Tests 

Predictive risk-based tests are the newest kind of testing available, and although they are seen more commonly in human genetic testing, they are new in veterinary genetic testing and will likely become commonplace in the future as technology, dog health, and behaviour datasets grow. Predictive risk-based tests are based on machine learning, sometimes called AI or “artificial intelligence,” which is constituted by an algorithm and a training dataset. The data used can vary, but should include a large number of dogs with genetic testing and paired medical, behavioral, or other records for best accuracy. An algorithm is developed by using all those pieces of data to determine which appears to predict or contribute to risk of the disorder, trait, or behaviour in question. Examples of data which may be used in the algorithm to predict risk include breed, breed type, sex, age, geography, care history, and dozens to thousands of genetic markers. The result is a predictive risk score for that individual, which can vary in predictive confidence, and is usually reported as their risk of developing the disorder, trait, or behaviour in their lifetime, compared to other dogs. Predictive risk-based tests are usually used in predicting complex disorders or behaviours that are known to be multifactorial in nature.