Read more about it below.

flybody – An Anatomically-Detailed Body Model of Drosophila melanogaster (Fruit Fly)

The AI model of a fruit fly that has been jointly developed by Google DeepMind and HHMI Janelia is called ‘flybody’ and it is an anatomically-detailed body model of the fruit fly – Drosophila melanogaster, and it has been made for both the MuJoCo physics simulator and also for reinforcement learning applications.

In its development process, Google DeepMind’s open-source physics simulator MuJoCo was used and features for simulating fluid forces on flapping wings to enable flight and adhesion actuators to mimic the gripping force of insect feet were added. Later an artificial neural network was trained with the help of recorded videos, and real fly behaviors were studied. Afterwards, this artificial neural network was made to control the fly model on MuJoCo, thereby letting the virtual insect move in a realistic manner – similar to how an actual real-world fly would move. Also, using paths marked by blue dots, the virtual insect model was also instructed to fly along complex natural flight trajectories as well.

Speaking of the use case of such AI models, these are expected to help scientists and researchers in understanding how an animal’s brain, body, and also its environment will drive specific behavioral patterns. To add, this has already been used to develop virtual models of multiple organisms – including a rodent and the above fruit fly. Next, it will be applied and used to develop the virtual model of a zebrafish.

The post Google DeepMind Partners With HHMI Janelia Research Campus to Develop an AI Model to Simulate a Fruit Fly appeared first on The Tech Outlook.

The AlphaFold 3 given the list of molecules can generate the joint 3D structure of it. It can also model chemical modifications to these molecules which control the healthy functioning of cells, that when disrupted can lead to disease. To perform such a complex task AlphaFold 3 after processing the input assembles its predictions using a diffusion network, similar to those found in AI image generators. It’s a crucial step to understand the structure of a particular drug and how it interacts with different molecules and accuracy is something that can’t be compromised in it.

Play video

Google does mention in its blog post not once but multiple times that this model achieves unprecedented accuracy in predicting drug-like interactions, including the binding of proteins with ligands and antibodies with their target proteins. Its sister company Isomorphic Labs is working on drug design for internal projects as well as with pharmaceutical partners. Scientists can access the majority of its capabilities, for free, through the newly launched AlphaFold Server.

The post Google DeepMind Reveals A New AI Model AlphaFold 3 Capable Of Accurately Predicting Protein Structure And Its Interaction With Molecules appeared first on The Tech Outlook.