Abstract
Ever-increasing numbers of quantitative proteomics data sets constitute an underexploited resource for investigating protein function. Multiprotein complexes often follow consistent trends in these experiments, which could provide insights about their biology. Yet, as more experiments are considered, a complex's signature may become conditional and less identifiable. Previously we successfully distinguished the general proteomic signature of genuine chromosomal proteins from hitchhikers using the Random Forests (RF) machine learning algorithm. Here we test whether small protein complexes can define distinguishable signatures of their own, despite the assumption that machine learning needs large training sets. We show, with simulated and real proteomics data, that RF can detect small protein complexes and relationships between them. We identify several complexes in quantitative proteomics results of wild-type and knockout mitotic chromosomes. Other proteins covary strongly with these complexes, suggesting novel functional links for later study. Integrating the RF analysis for several complexes reveals known interdependences among kinetochore subunits and a novel dependence between the inner kinetochore and condensin. Ribosomal proteins, although identified, remained independent of kinetochore subcomplexes. Together these results show that this complex-oriented RF (NanoRF) approach can integrate proteomics data to uncover subtle protein relationships. Our NanoRF pipeline is available online.