Studying the brain mechanisms behind consciousness is a major challenge for neuroscience and medicine. Accumulating evidence shows that the structural, histological, functional, genetic, and neurochemical inhomogeneities of the mammalian cortex do not follow a modular distribution; instead, these properties change following gradients, understood as axes of variance along which cortical features are ordered continuously. The gradient describing the axis of largest variance (principal gradient) obtained for an ample range of cortical features follows a unimodal-transmodal organization, ranging from externally-oriented sensory and motor regions to multimodal association regions, culminating in regions linked with internally oriented higher-order cognitive functions. In this project we propose a novel approach, constructing, validating and exploring whole-brain computational models combining empirical information including anatomical connectivity, spatial maps of local neuroanatomical features, to reproduce the configuration of human functional gradients, as determined using manifold learning techniques applied to functional magnetic resonance imaging (fMRI) data. This will allow us to investigate the process by which functional gradients emerge from the spatial distribution of cortical anatomical inhomogeneities. The models will also provide the possibility to investigate how different global brain states behave under perturbations. In order to achieve our goals, we propose a highly interdisciplinary project that combines state-of-the-art principal gradient expertise with whole-brain computational modelling proposing a synergy between two groups with large expertise in each area to address a common question: do realistic functional gradients emerge from the dynamical equations when coupled by realistic long-range structural connections, and modulated locally by empirical maps encoding relevant neurochemical data?