Incorporating support vector machine with sequential minimal optimization to identify anticancer peptides

Abstract Background Cancer is one of the major causes death worldwide. To treat cancer, use anticancer peptides (ACPs) has attracted increased attention in recent years. ACPs are a unique group small molecules that can target and kill cancer cells fast directly. However, identifying by wet-lab experiments time-consuming labor-intensive. Therefore, it significant to develop computational tools for prediction. Though some ACP prediction have been developed recently, their performances not well enough most them do offer function distinguish from antimicrobial (AMPs). Considering fact growing number studies shown AMPs exhibit function, this work tries build model distinguishing addition predicts whole peptides. Results This study chooses amino acid composition, N5C5, k-space, position-specific scoring matrix (PSSM) as features, analyzes machine learning methods, including support vector (SVM) sequential minimal optimization (SMO) (model 2) Another 1) distinguishes also developed. Comparing previous models, models research show better performance (accuracy: 85.5% 1 95.2% 2). Conclusions utilizes new feature, PSSM, which contributes than other features. In SVM, SMO used optimizing SVM SMO-optimized non-optimized models. Last but least, provides two different functions, all The second model, PSSM performs existing tools.