Exploration of drug–target interactions (DTIs) requirement a financial, human and materialistic resources in conducting biomedical experimentations. In order to reduce the cost and time to meet the present needs Artificial Intelligence (AI) is introduced that helps in predicting the DTIs. With available target and drug data in conventional databases enables the machine or deep learning model a mainstream technology for DTIs. In this paper, we develop anImproved Frequent Subsequence Mining(IFSM) based transformer binding phase (TBP) for the pre-processing and extraction of features for drug-target interaction. At the initial phase, we use IFSM to extract the meaningful frequent subsequence from the input datasets that forms an intuitive pattern expression. The study aims at extraction of the instances by initially transforming the datasets to thesub-structures using Improved Frequent Subsequence Mining (IFSM).The TBP enables the extraction of semantic relations between the sub-structures extracted from previous IFSM phase, where it is of an unlabelled biomedical data. The simulation is conducted to test the efficacy of the model is tested on state-of-art DTI feature extraction models to test the efficacy of accurate contextual structural binding generation. The efficacy of the model is tested in terms of accuracy, precision, recall and f-measure.

Drug–Target Interactions, Improved Frequent Subsequence Mining, Transformer Binding Phase