Networks represent now an important part of post‐genomic data. This gives rise to a wide variety of statistical problems involving random graphs. In this talk, I will present recent advances made in my "Statistics for Systems Biology" group (www.ssbgroup.fr). Namely, I will present (i) how to infer gene regulatory networks from gene expression data, (ii) how to model biological networks thanks to a mixture of random graphs (MixNet/Stochastic Block Model) and to recover the latent structure, eventually allowing nodes to belong to more than one group, (iii) how to assess the significance of network motif frequencies in an observed network.