Scientific work now facing the needs on dynamic scientific data. Scientific data need to be interactive, reproducible, collaborative, and dynamic and have the reputation and influences. The common collaborative framework such as cloud and semantic technologies has being give influences to the scientific data. The development of new framework is compete with the needs of multidisciplinary research, the exploding of digital universe, the industry needs on long-term use and access, and some issues in next generation data, such as data integration, annotation, provenance and security. In this research, we introduced Science-Forge, the scientific collaboration framework that addresses the issues in dynamic scientific data and tools. The framework inspired by GForge system that enabled software developer to share and collaborate in software development. We build the scientific preservation based on GForge system, and implementing several data process. The framework has capabilities to handling data collection, data processing. The research aimed is to establishment of the new framework for scientific collaboration works. ©2010 IEEE.