With new generation sequencing, software tools are essential to efficiently integrate molecular biological data around phenotypic concepts. The purpose of this work was to enhance the access to proteins from diseases. The first step consisted in creating an automatic tool to link disease-related human proteins and variants to a medical controlled vocabulary and make this mapping accessible through a web interface. Then another tool was proposed to filter networks of protein-protein interactions with single clinical traits (phenotypes) found in Mendelian disorders. The aim was to help formulating hypotheses on the function of proteins and interactions by isolating a spatio-temporal context, crucial given the modular nature of cell biology.