AI2BMD: A Quantum-Precise Machine Learning Approach for Large-Scale Biomolecular Dynamics

Biomolecular dynamics simulations are crucial for biological sciences and provide insights into molecular interactions. Although classical molecular dynamics (MD) simulations are efficient, they lack chemical precision. Methods such as density functional theory (DFT) achieve high precision, but are too computationally intensive for large biomolecules. MD simulations allow the observation of molecular behavior: classical MD uses interatomic potentials and ab initio MD (AIMD) derives forces from electronic structures. AIMD's scalability issues limit its use in biomolecular studies. Machine learning force fields (MLFFs), trained with DFT-level data, promise accuracy at lower costs, although generalization across diverse molecular conformations remains a challenge.

Researchers at Microsoft Research in Beijing introduced AI2BMD, an artificial intelligence-based system for simulating large biomolecules with ab initio precision. AI2BMD uses a protein fragmentation technique and a machine learning force field, allowing it to accurately calculate the energy and forces of proteins with more than 10,000 atoms. This system is much more efficient than traditional DFT and reduces simulation times by orders of magnitude. AI2BMD can perform hundreds of nanoseconds of simulations, capturing protein folding, unfolding, and conformational dynamics. Its thermodynamic predictions align closely with experimental data, making it a valuable tool to complement wet lab experiments and advance biomedical research.

The protein fragmentation approach is based on the fundamental structure of amino acids in proteins, where each amino acid contains a main chain of atoms (Cα, C, O, N and H) and a distinct side chain. To create a model that is widely applicable to various proteins, each amino acid is treated as a dipeptide, capped with Ace and Nme groups at its ends. This approach, based on overlapping dipeptide fragments, helps ensure comprehensive protein coverage. Using a sliding window, protein chains are divided into these dipeptides, where each fragment includes atoms of the main chain and partial atoms of adjacent amino acids. This method accurately calculates protein energies and atomic forces by adding hydrogens as needed to the Cα bonds and optimizing the positions using a quasi-Newton algorithm. This generalizable method allows systematic application to all proteins, reducing complexities and maximizing model accuracy.

The training data set for the potential AI2BMD involves sampling millions of dipeptide conformations to capture the variety of protein structures. A deep learning model called ViSNet was trained using this extensive data set to predict atomic energy and forces based on atomic numbers and coordinates. The model used specific hyperparameters to optimize accuracy and was trained with early stopping techniques. Simulations based on AI2BMD's potential are processed by a cloud-compatible ai-powered simulation program, allowing flexible deployment across computing environments. This system supports parallelized simulation processes and automatically preserves progress in cloud storage, ensuring robust and efficient handling of protein dynamics modeling.

AI2BMD shows significant potential in protein property estimation, especially for thermodynamic analysis of rapidly folding proteins. AI2BMD could classify structures into folded and unfolded states by simulating various types of proteins and accurately predicting potential energy values. Their melting temperature (Tm) estimates for proteins such as the WW domain and NTL9 were in close agreement with experimental data, often outperforming traditional molecular mechanics (MM) methods. Furthermore, AI2BMD calculations for free energy (ΔG), enthalpy, and heat capacity were highly consistent with experimental findings, reinforcing their accuracy. This robustness in thermodynamic estimation highlights the value of AI2BMD as an advanced tool for protein analysis.

In addition to thermodynamics, AI2BMD was shown to be effective in alchemical free energy calculations, such as pKa prediction, and is valuable in biochemical research. Unlike traditional QM-MM methods that restrict calculations to preset regions, AI2BMD's ab initio approach allows modeling of entire proteins without boundary inconsistencies, making it versatile for complex proteins and dynamic states. Although the speed of AI2BMD is still slower than that of classical MD, future optimizations and applications to other biomolecular systems could improve its efficiency. The adaptability of AI2BMD makes it a promising tool for drug discovery, protein design, and enzyme engineering, offering high-precision simulations for various biomolecular applications.

look at the Paper and Details. All credit for this research goes to the researchers of this project. Also, don't forget to follow us on twitter.com/Marktechpost”>twitter and join our Telegram channel and LinkedIn Grabove. If you like our work, you will love our information sheet.. Don't forget to join our SubReddit over 55,000ml.

(Sponsorship opportunity with us) Promote your research/product/webinar to over 1 million monthly readers and over 500,000 community members

Sana Hassan, a consulting intern at Marktechpost and a dual degree student at IIT Madras, is passionate about applying technology and artificial intelligence to address real-world challenges. With a strong interest in solving practical problems, he brings a new perspective to the intersection of ai and real-life solutions.