The study presents a computational model of the *Drosophila melanogaster* brain, based on the recently assembled central brain connectome, which contains over 125,000 neurons and 50 million synaptic connections. The model uses a leaky integrate-and-fire approach to simulate neural circuits and predict sensory processing and motor behaviors. Specifically, the model is used to study feeding and grooming behaviors, focusing on the interaction of taste modalities and mechanosensation. In the feeding initiation circuit, the model accurately predicts neurons that respond to sugar and water tastes, as well as those required for proboscis extension. For example, activating sugar-sensing neurons in the model predicts the activation of specific motor neurons (MNs) involved in feeding. The model also correctly identifies neurons that are necessary for feeding initiation, such as MN9, which controls rostrum lifting. Additionally, the model predicts the interaction between different taste modalities, such as sugar and bitter, and the role of Ir94e neurons, which respond to salt and male genitalia but are found to inhibit proboscis extension. In the antennal grooming circuit, the model predicts the activation of key interneurons and descending neurons involved in antennal grooming. It identifies neurons that are sufficient for grooming and those that are required for it. The model also correctly predicts the response of specific neurons to different mechanosensory stimuli, such as JO-CE and JO-F neurons. Overall, the computational model demonstrates that a simple, connectivity-based approach can accurately predict complex sensorimotor transformations in the fly brain, providing a powerful tool for studying sensory processing and neural circuit functions.The study presents a computational model of the *Drosophila melanogaster* brain, based on the recently assembled central brain connectome, which contains over 125,000 neurons and 50 million synaptic connections. The model uses a leaky integrate-and-fire approach to simulate neural circuits and predict sensory processing and motor behaviors. Specifically, the model is used to study feeding and grooming behaviors, focusing on the interaction of taste modalities and mechanosensation. In the feeding initiation circuit, the model accurately predicts neurons that respond to sugar and water tastes, as well as those required for proboscis extension. For example, activating sugar-sensing neurons in the model predicts the activation of specific motor neurons (MNs) involved in feeding. The model also correctly identifies neurons that are necessary for feeding initiation, such as MN9, which controls rostrum lifting. Additionally, the model predicts the interaction between different taste modalities, such as sugar and bitter, and the role of Ir94e neurons, which respond to salt and male genitalia but are found to inhibit proboscis extension. In the antennal grooming circuit, the model predicts the activation of key interneurons and descending neurons involved in antennal grooming. It identifies neurons that are sufficient for grooming and those that are required for it. The model also correctly predicts the response of specific neurons to different mechanosensory stimuli, such as JO-CE and JO-F neurons. Overall, the computational model demonstrates that a simple, connectivity-based approach can accurately predict complex sensorimotor transformations in the fly brain, providing a powerful tool for studying sensory processing and neural circuit functions.