The worldwide appearance of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has generated significant apprehension and poses a considerable obstacle to global health. As a common post-translational modification, phosphorylation impacts many vital cellular functions and is intimately connected to SARS-CoV-2 infection. Therefore, precisely identifying phosphorylation sites could provide a more in-depth insight into the processes behind SARS-CoV-2 infection and help alleviate the continuing COVID-19 crisis. However, currently available computational prediction tools for predicting these sites are lacking in accuracy and effectiveness. In this study, we designed an innovative meta-learning model, MeL-STYPhos (Meta-Learning for Serine/Threonine and Tyrosine Phosphorylation), to precisely identify protein phosphorylation sites. We initially performed a comprehensive assessment of 29 unique sequence-derived features, establishing prediction models for each using 14 renowned machine learning methods, ranging from traditional classifiers to advanced deep learning algorithms. We then selected the most effective model for each feature encoding technique, integrating their predicted values. Rigorous feature selection strategies were employed to identify the optimal base models and classifier for each phosphorylation site. To our knowledge, this represents the first study to utilize such an extensive range of sequence-derived features and machine learning algorithms for phosphorylation site prediction, ultimately leading to the development of MeL-STYPhos. Extensive cross-validation and independent testing revealed that our method surpasses existing state-of-the-art tools in phosphorylation site prediction. We have also developed a publicly accessible platform at https://balalab-skku.org/MeL-STYPhos. We believe that MeL-STYPhos will serve as a valuable tool for accelerating the discovery of S/T/Y phosphorylation events and elucidating their role in post-translational regulation.