Quantitative Structure‐Fluorescence Property Relationship Analysis of a Large BODIPY Library

Schüller, Andreas; Goh, Garrett B.; Kim, Hanjo; Lee, Jung Hoon; Chang, Young‐Tae

doi:10.1002/minf.201000089

Cited by 23 publications

(28 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While DNN applications in material design is still at its infancy, it would be interesting to see how its application will fare against traditional QSPR applications and upcoming rational materials design endeavors, such as in the prediction of spectral properties of fluorophores, properties of ionic liquids, and nanostructure activity …”

Section: Computational Materials Designmentioning

confidence: 99%

Deep learning for computational chemistry

2017

View full text Add to dashboard Cite

The rise and fall of artificial neural networks is well documented in the scientific literature of both computer science and computational chemistry. Yet almost two decades later, we are now seeing a resurgence of interest in deep learning, a machine learning algorithm based on multilayer neural networks. Within the last few years, we have seen the transformative impact of deep learning in many domains, particularly in speech recognition and computer vision, to the extent that the majority of expert practitioners in those field are now regularly eschewing prior established models in favor of deep learning models. In this review, we provide an introductory overview into the theory of deep neural networks and their unique properties that distinguish them from traditional machine learning algorithms used in cheminformatics. By providing an overview of the variety of emerging applications of deep neural networks, we highlight its ubiquity and broad applicability to a wide range of challenges in the field, including QSAR, virtual screening, protein structure prediction, quantum chemistry, materials design and property prediction. In reviewing the performance of deep neural networks, we observed a consistent outperformance against nonneural networks state-of-the-art models across disparate research topics, and deep neural network based models often exceeded the "glass ceiling" expectations of their respective tasks. Coupled with the maturity of GPU-accelerated computing for training deep neural networks and the exponential growth of chemical data on which to train these networks on, we anticipate that deep learning algorithms will be a valuable tool for computational chemistry. 3 IntroductionDeep Learning is the key algorithm used in the development of AlphaGo, a Deep learning is a machine learning algorithm, not unlike those already in use in various applications in computational chemistry, from computer-aided drug design to materials property prediction. 5-8 Amongst some of its more high profile achievements include the Merck activity prediction challenge in 2012, where a deep neural network not only won the competition and outperformed Merck's internal baseline model, but did so without having a single chemist or biologist in their team. In a repeated success by a different research team, deep learning models achieved top positions in the Tox21 toxicity prediction challenge issued by NIH in 2014. 9 The unusually stellar performance of deep learning models in both predicting activity and toxicity in these recent examples, originate from the unique characteristics that distinguishes deep learning from traditional machine learning algorithms.For those unfamiliar with the intricacies of machine learning algorithms, we will highlight some of the key differences between traditional (shallow) machine learning and deep learning.The simplest example of a machine learning algorithm would be the ubiquitous least-squares linear regression. In linear regression, the underlying nature of the model is known (linear in th...

show abstract

Section: Computational Materials Designmentioning

confidence: 99%

Deep learning for computational chemistry

2017

View full text Add to dashboard Cite

show abstract

“…Whereas in principle quantum mechanical methods are efficacious as long as the physical approximations remains reasonable, empirical models such as QSAR and ML typically relies heavily on the scope of the training set and thereby lacks universality. For example, the published QSAR studies on the relationship between molecular structures and photophysical properties are usually limited to a maximum of hundreds of molecules 72,73 . It is therefore important to assess the scope of our ML model (hence the potential in real-world applications) for the prediction of emission wavelengths.…”

Section: Machine Learning Predictions For Plqymentioning

confidence: 99%

Machine Learning Enables Highly Accurate Predictions of Photophysical Properties of Organic Fluorescent Materials: Emission Wavelengths and Quantum Yields

Bai²,

Li³

et al. 2020

Preprint

View full text Add to dashboard Cite

<div> <p>The development of functional organic fluorescent materials calls for fast and accurate predictions of photophysical parameters for processes such as high-throughput virtual screening, while the task is challenged by the limitations of quantum mechanical calculations. We establish a database covering >4,300 solvated organic fluorescent dyes and develop new machine learning (ML) approach aimed at efficient and accurate predictions of emission wavelength and photoluminescence quantum yield (PLQY). Our feature engineering has given rise to Functionalized Structure Descriptor (FSD) and Comprehensive General Solvent Descriptor (CGSD), whereby a highly black-box computational framework is realized with consistently good accuracy across different dye families, ability of describing substitution effects and solvent effects, efficiency for large-scale predictions and workability with on-the-fly learning. Evaluations with unseen molecules suggests a remarkable MAE of 0.13 for PLQY and 0.080 eV for emission energy, the latter comparable to time-dependent density functional theory (TD-DFT) calculations. An online prediction platform was constructed based on the ensemble model to make prediction in various solvents (https://www.chemfluor.top/). Our statistical learning methodology will complement quantum mechanical calculations as an efficient alternative approach for the prediction of these parameters.<br></p> </div><p> <br></p>

show abstract

“…A number of previous studies investigated quantitative AD: (1) range setting of each variable or latent variable after dimensional reduction [21] by Principal Component Analysis (PCA) [22], (2) distance from neighboring point [23] using kNN [20], (3) data density estimation [24] by OneClassSVM (OCSVM) [25], and (4) variation of predicted values [26] by ensemble learning [8,9]. [2,3,4,5,6,7,8,9,10,15,20] In this study, we proposed to use Tanimoto distance of binary unhashed Morgan fingerprint (radius = 4) with the training data as the AD estimation model. First, Tanimoto distances were calculated for each compound against all compounds in the training data.…”

Section: Applicability Domain (Ad)mentioning

confidence: 99%

“…Among the fluorescent substances, those having boron-dipyrromethene (BODIPY) in the molecular skeleton have a sharp spectrum [3] and stably high quantum yield [4], and the influence of solvent to the wavelength is small. Schüller et al [5] constructed a QSPR model for compounds with BODIPY skeleton by combining multiple variable selection methods.…”

Section: Introductionmentioning

confidence: 99%

Ensemble Machine Learning and Applicability Domain Estimation for Fluorescence Properties and its Application to Structural Design

Sugawara

Kotera

Tanaka

et al. 2019

J. Comput. Aided Chem.

View full text Add to dashboard Cite

Fluorescent substances are used in a wide range of applications, and the method that effectively design molecules having desirable absorption and emission wavelength is required. In this study, we used borondipyrromethene (BODIPY) compounds as a case study, and constructed high precision wavelength prediction model using ensemble learning. Prediction accuracy improved in stacking model using RDKit descriptors and Morgan fingerprint. The variables related to the molecular skeleton and the conjugation length were shown to be important. We also proposed an applicability domain (AD) estimation model that directly use the descriptors based on Tanimoto distance. The performance of the AD models was shown better than the OCSVM-based model. Using our proposed stacking model and AD model, newly generated compounds were screened and we obtained 602 compounds which were estimated inside the AD in both absorption wavelength and emission wavelength.

show abstract

Quantitative Structure‐Fluorescence Property Relationship Analysis of a Large BODIPY Library

Cited by 23 publications

References 62 publications

Deep learning for computational chemistry

Deep learning for computational chemistry

Machine Learning Enables Highly Accurate Predictions of Photophysical Properties of Organic Fluorescent Materials: Emission Wavelengths and Quantum Yields

Ensemble Machine Learning and Applicability Domain Estimation for Fluorescence Properties and its Application to Structural Design

Contact Info

Product

Resources

About