Coronavirus disease 2019 (COVID-19) is a contagious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). This disease has spread globally, causing more than 161.5 million cases and 3.3 million deaths. Keeping on the identification, surveillance and the study of the temporal dynamics of mutations with significant representation is central to understand the adaptation of SARS-CoV-2. Furthermore, how lockdown policies influence the dynamics of SARS-CoV-2 mutations is poorly understood. Here, using 1 058 020 SARS-CoV-2 genomes and COVID-19 cases from 714 country-month combinations representing 98 countries, we performed a normalization by COVID-19 cases calculation of relative frequency of SARS-CoV-2 mutations. We found 115 mutations estimated to be present in more than 3 % of global COVID-19 cases and determined three types of mutation dynamics: High-Frequency, Medium-Frequency, and Low-Frequency. Classification of mutations based on temporal dynamics helps to study viral adaptation and can be used to evaluate the effects of human behaviors in the pandemic. For instance, we report a negative and positive correlation of the frequency change of High-Frequency mutations with the level of international movement controls and the number of flight departures, respectively.