Algorithm A1 Drug Recommendation Model Algorithm for Greedy Strategies

Require: current_drugs, combo_freq_dict, risk_dict, gene_map, α, β, N Ensure: Top_N_Recommendations  1:  Begin  2:  Normalize current_drugs  3:  all_drugs ← union of all drugs in combo_freq_dict  4: candidate_drugs ← all_drugs − current_drugs  5: scored_candidates ← empty list  6:  for each drug_c in candidate_drugs do  7:    extended_combo ← current_drugs ∪ {drug_c}  8:    freq_score ← 0  9:    for each subset in combinations of extended_combo with size 2 to 3 do 10:       freq_score ← freq_score + combo_freq_dict.get(subset, 0) 11:   end for 12:   risk_score ← 0 13:   for each drug_d in current_drugs do 14:      pair ← {drug_c, drug_d} 15:      risk ← risk_dict.get(pair, “Unknown”) 16:      risk_score ← risk_score + penalty_value(risk) 17:   end for 18:   overlap_score ← 0 19:   for each drug_d in current_drugs do 20:      overlap_score ← overlap_score + size of (gene_map[drug_c] ∩ gene_map[drug_d]) 21:   end for 22:   total_score ← α× freq_score + β × overlap_score + risk_score 23:   Append (drug_c, total_score) to scored_candidates 24: end for 25: Sort scored_candidates in descending order of total_score 26: Top_N_Recommendations ← top N elements from scored_candidates 27: End