Improved Candidate Drug Mining for Alzheimer’s Disease

Algorithm 2

Pseudocode for a priori algorithm for the data mining in PharmGKB, where 𝜀 is a support threshold, 𝐿 is the frequent gene subsets that satisfy the support threshold, 𝑘 is the number of current iterations, and 𝐶 is the candidate set, and count[gene] accesses a field of the data structure that represents gene candidate set.

(1) Apriori(PharmGKB, 𝜀)

(2)  𝐿₁← (frequent genes in drug class for Alzheimer’s disease)

(3)  𝑘←2

(4) while 𝐿_𝑘−1≠𝜙

(5)  𝐶_𝑘←𝑎∪{𝑏}∣𝑎∈𝐿_𝑘−1∧𝑏∈⋃𝐿_𝑘−1∧𝑏∉𝑎

(6)  for each drug class ∈ PharmGKB

(7)   𝐶_𝑡←gene∣gene∈𝐶_𝑘∧gene⊆drugclass

(8)   for each candidate gene ∈ 𝐶_𝑡

(9)    count[gene] ← count[gene] + 1

(10)    end for

(11)  end for

(12)  𝐿_𝑘←gene∣gene∈𝐶_𝑘∧count[gene]>𝜀

(13)  𝑘←𝑘+1

(14) end while

(15) return ⋃_𝑘𝐿_𝑘