A novel machine learning method for cytokine-receptor interaction prediction.

In this study, we propose a novel machine learning based method for cytokine-receptor interaction prediction. In the proposed method, a protein sequence is first transformed by incorporating the sequence evolutional information and then formulated by the following three aspects: (1) the k-skip-n-gram model, (2) physicochemical properties, and (3) local pseudo position-specific score matrix (local PsePSSM). Afterwards, Random Forest (RF) classifier is employed to make the predictions for potential cytokine-receptor interactions. Experimental results on a dataset from Homo sapiens show that the proposed method performs remarkably better than existing methods, leading by 3.4% for the overall prediction accuracy. PMID: 26552440 [PubMed - as supplied by publisher]
Source: Combinatorial Chemistry and High Throughput Screening - Category: Chemistry Authors: Tags: Comb Chem High Throughput Screen Source Type: research