In many complex biological processes quantitative data is scarce, which makes it problematic to create accurate quantitative models of the system under study. In this work, we suggest that the Qualitative Networks (QNs) framework is an appropriate approach for modeling biological networks when only little quantitative data is available. Using QNs we model a metabolic network related to fat metabolism, which plays an important role in type-2 diabetes and obesity. The model is based on gene expression data of the regulatory network of a key transcription factor Mlxipl. Our model reproduces the experimental data and allows in-silico testing of new hypotheses. Specifically, the QN framework allows to predict new modes of interactions between components within the network. Furthermore, we demonstrate the value of the QNs approach in directing future experiments and its potential to facilitate our understanding of the modeled system.