Weighted sequence motifs as an improved seeding step in microRNA target prediction algorithms

摘要: We present a new microRNA target prediction algorithm called TargetBoost, and show that the is stable identifies more true targets than do existing algorithms. TargetBoost uses machine learning on set of validated in lower organisms to create weighted sequence motifs capture binding characteristics between microRNAs their targets. Existing algorithms require candidates have (1) near-perfect complementarity microRNAs' 5' end targets; (2) relatively high thermodynamic duplex stability; (3) multiple sites target's 3' UTR; (4) evolutionary conservation species. Most use one two first requirements seeding step, three others as filters improve method's specificity. The initial step determines an algorithm's sensitivity also influences its As all may add increase specificity, we propose methods should be compared before such filtering. TargetBoost's motif approach favorable using both stability steps. (TargetBoost available Web tool from http://www.interagon.com/demo/.).