A5026256128- Machine Learning in Materials Science
- Computational Drug Discovery Methods
- Various Chemistry Research Topics
- Radioactive element chemistry and processing
- GaN-based semiconductor devices and materials
- Ga2O3 and related materials
- Genetics, Bioinformatics, and Biomedical Research
- Chemistry and Chemical Engineering
- Protein Structure and Dynamics
- Machine Learning and Algorithms
- Metal and Thin Film Mechanics
- History and advancements in chemistry
- Data Quality and Management
- Analytical Chemistry and Chromatography
- Crystallization and Solubility Studies
- Spectroscopy Techniques in Biomedical and Chemical Research
- Experimental Learning in Engineering
- Acoustic Wave Resonator Technologies
- Distributed systems and fault tolerance
- Scientific Computing and Data Management
- Engineering Education and Pedagogy
- Biomedical and Engineering Education
- Electronic and Structural Properties of Oxides
Massachusetts Institute of Technology
2020-2025
Moscow Institute of Thermal Technology
2023
University of Michigan
2019
Deep learning has become a powerful and frequently employed tool for the prediction of molecular properties, thus creating need open-source versatile software solutions that can be operated by nonexperts. Among current approaches, directed message-passing neural networks (D-MPNNs) have proven to perform well on variety property tasks. The package Chemprop implements D-MPNN architecture offers simple, easy, fast access machine-learned properties. Compared its initial version, we present...
A closed-loop, autonomous molecular discovery platform driven by integrated machine learning tools was developed to accelerate the design of molecules with desired properties. We demonstrated two case studies on dye-like molecules, targeting absorption wavelength, lipophilicity, and photooxidative stability. In first study, experimentally realized 294 unreported across three automatic iterations design-make-test-analyze cycles while exploring structure-function space four rarely reported...
ConspectusDesigning new materials is vital for addressing pressing societal challenges in health, energy, and sustainability. The combination of physicochemical laws empirical trial error has long guided material design, but this approach limited by the cost experiments difficulty deriving complex guiding principles. space hypothetical to be considered incredibly large, only a small fraction possible compounds can ever tested experimentally. computational techniques atomistic simulation...
Optical properties are central to molecular design for many applications, including solar cells and biomedical imaging. A variety of
Machine learning sequence-function models for proteins could enable significant advances in protein engineering, especially when paired with state-of-the-art methods to select new sequences property optimization and/or model improvement. Such (Bayesian and active learning) require calibrated estimations of uncertainty. While studies have benchmarked a variety deep uncertainty quantification (UQ) on standard molecular machine-learning datasets, it is not clear if these results extend...
Deep learning has become a powerful and frequently employed tool for the prediction of molecular properties, thus creating need open-source versatile software solutions that can be operated by non-experts. Among current approaches, directed message-passing neural networks (D-MPNNs) have proven to perform well on variety property tasks. The package Chemprop implements D-MPNN architecture, offers simple, easy, fast access machine-learned properties. Compared its initial version, we present...
Deep learning has become a powerful and frequently employed tool for the prediction of molecular properties, thus creating need open-source versatile software solutions that can be operated by non-experts. Among current approaches, directed message-passing neural networks (D-MPNNs) have proven to perform well on variety property tasks. The package Chemprop implements D-MPNN architecture, offers simple, easy, fast access machine-learned properties. Compared its initial version, we present...
Discovery of high-performance materials and molecules requires identifying extremes with property values that fall outside the known distribution. Therefore, ability to extrapolate out-of-distribution (OOD) is critical for both solid-state molecular design. Our objective train predictor models zero-shot higher ranges than in training data, given chemical compositions solids or graphs their values. We propose using a transductive approach OOD prediction, achieving improvements prediction...
A closed-loop, autonomous molecular discovery platform driven by integrated machine learning tools was developed to accelerate the design of molecules with desired properties. Two case studies are demonstrated on dye-like molecules, targeting absorption wavelength, lipophilicity, and photo-oxidative stability. In first, experimentally realized 312 unreported across three automatic iterations design-make-test-analyze cycles while exploring structure–function space four rarely reported...
Deep learning has become a powerful and frequently employed tool for the prediction of molecular properties, thus creating need open-source versatile software solutions that can be operated by non-experts. Among current approaches, directed message-passing neural networks (D-MPNNs) have proven to perform well on variety property tasks. The package Chemprop implements D-MPNN architecture, offers simple, easy, fast access machine-learned properties. Compared its initial version, we present...
Abstract Machine learning sequence-function models for proteins could enable significant ad vances in protein engineering, especially when paired with state-of-the-art methods to select new sequences property optimization and/or model improvement. Such (Bayesian and active learning) require calibrated estimations of uncertainty. While studies have benchmarked a variety deep uncertainty quantification (UQ) on standard molecular machine-learning datasets, it is not clear if these results...
Automated patent mining creates domain-specific datasets of molecular structures for generative modeling with limited human intervention.
Organic optoelectronic materials are a promising avenue for next-generation electronic devices due to their solution processability, mechanical flexibility, and tunable properties. In particular, near-infrared (NIR) sensitive molecules have unique applications in night-vision equipment biomedical imaging. Molecular engineering has played crucial role developing non-fullerene acceptors (NFAs) such as the Y-series molecules, which significantly improved power conversion efficiency (PCE) of...
Optical properties are central to molecular design for many applications, including solar cells and biomedical imaging. A variety of ab initio statistical methods have been developed their prediction, each with a trade-off between accuracy, generality, cost. Existing theoretical such as time-dependent density functional theory (TD-DFT) generalizable across chemical space because robust physics-based foundations but still exhibit random systematic errors respect experiment despite high...
InGaN light-emitting diodes (LEDs) are more efficient and cost effective than incandescent fluorescent lighting, but lattice mismatch limits the thickness of layers that can be grown on GaN without performance-degrading dislocations. In this work, we apply hybrid density functional theory calculations to investigate thermodynamic stability, parameters, bandgaps wurtzite zincblende quaternary BInGaN alloys. We find phase is stable matched for compositions containing up ∼30% boron. The match...