Trinucleotide Repeat (TNR) Expansions Mining Algorithm Based on the Apriori Algorithm

The Apriori algorithm is a data mining algorithm used exclusively for identifying associations between itemsets in a series of business transactions. In this paper, we modify this algorithm to identify associations between trinucleotide repeat expansions in DNA, which are known for their associations with specific genetic diseases. The proposed modification here is to treat all possible forms of DNA trinucleotides as item sets and the list of all DNA segments as a series of transactions. In the first iteration of the algorithm, the number of occurrences of every single trinucleotide is counted across all DNA segments. Then, the results are filtered out according to a preset threshold. In the second iteration, the number of occurrences of every pair of trinucleotides is counted, and similarly, the results are filtered out according to a threshold. In the third iteration, the number of occurrences of every trio of trinucleotides is handled similarly. The algorithm continues until no results are produced. Lastly, the association rules are formed, yielding the conclusion of the existence of a certain genetic disease will also mean an existence of another genetic disease.