Understanding how biological systems develop and function remains one of the main open scientific challenges of our times. An improved quantitative understanding of biological systems, assisted by computational models is also important for future bioengineering and biomedical applications. We present a computational approach aimed towards unifying hypotheses with models and experiments, allowing to formally represent what a biological system does (specification) how it does it (mechanism) and systematically compare to data characterizing system behavior (experiments). We describe our Biocharts framework geared towards supporting this approach and illustrate its application in several biological domains including bacterial colony growth, developmental biology, and stem cell population dynamics.