AI-Powered Knowledge Base Enables Transparent Prediction of Nanozyme Multiple Catalytic Activity.

Nanozymes are unique materials with many valuable properties for applications in biomedicine, biosensing, environmental monitoring, and beyond. In this work, we developed a machine learning (ML) approach to search for new nanozymes and deployed a web platform, DiZyme, featuring a state-of-the-art database of nanozymes containing 1210 experimental samples, catalytic activity prediction, and DiZyme Assistant interface powered by a large language model (LLM). For the first time, we enable the prediction of multiple catalytic activities of nanozymes by training an ensemble learning algorithm achieving R2 = 0.75 for the Michaelis-Menten constant and R2 = 0.77 for the maximum velocity on unseen test data. We envision an accurate prediction of multiple catalytic activities (peroxidase, oxidase, and catalase) promoting novel applications for a wide range of surface-modified inorganic nanozymes. The DiZyme Assistant based on the ChatGPT model provides users with supporting information on experimental samples, such as synthesis procedures, measurement protocols, etc. DiZyme (dizyme.aicidlab.itmo.ru) is now openly available worldwide.

Machine Learning Reinforced Genetic Algorithm for Massive Targeted Discovery of Selectively Cytotoxic Inorganic Nanoparticles.

Nanoparticles (NPs) have been employed as drug delivery systems (DDSs) for several decades, primarily as passive carriers, with limited selectivity. However, recent publications have shed light on the emerging phenomenon of NPs exhibiting selective cytotoxicity against cancer cell lines, attributable to distinct metabolic disparities between healthy and pathological cells. This study revisits the concept of NPs selective cytotoxicity, and for the first time proposes a high-throughput in silico screening approach to massive targeted discovery of selectively cytotoxic inorganic NPs. In the first step, this work trains a gradient boosting regression model to predict viability of NP-treated cell lines. The model achieves mean cross-validation (CV) Q2 = 0.80 and root mean square error (RMSE) of 13.6. In the second step, this work develops a machine learning (ML) reinforced genetic algorithm (GA), capable of screening >14 900 candidates/min, to identify the best-performing selectively cytotoxic NPs. As proof-of-concept, DDS candidates for the treatment of liver cancer are screened on HepG2 and hepatocytes cell lines resulting in Ag NPs with selective toxicity score of 42%. This approach opens the door for clinical translation of NPs, expanding their therapeutic application to a wider range of chemical space of NPs and living organisms such as bacteria and fungi.

Quantifying the Efficacy of Magnetic Nanoparticles for MRI and Hyperthermia Applications via Machine Learning Methods.

Magnetic nanoparticles are a prospective class of materials for use in biomedicine as agents for magnetic resonance imagining (MRI) and hyperthermia treatment. However, synthesis of nanoparticles with high efficacy is resource-intensive experimental work. In turn, the use of machine learning (ML) methods is becoming useful in materials design and serves as a great approach to designing nanomagnets for biomedicine. In this work, for the first time, an ML-based approach is developed for the prediction of main parameters of material efficacy, i.e., specific absorption rate (SAR) for hyperthermia and r1 /r2 relaxivities in MRI, with parameters of nanoparticles as well as experimental conditions as descriptors. For that, a unique database with more than 980 magnetic nanoparticles collected from scientific articles is assembled. Using this data, several tree-based ensemble models are trained to predict SAR, r1 and r2 relaxivity. After hyperparameter optimization, models reach performances of R2 = 0.86, R2 = 0.78, and R2 = 0.75, respectively. Testing the models on samples unseen during the training shows no performance drops. Finally, DiMag, an open access resource created to guide synthesis of novel nanosized magnets for MRI and hyperthermia treatment with machine learning and boost development of new biomedical agents, is developed.

Regularized adversarial learning for normalization of multi-batch untargeted metabolomics data.

MOTIVATION: Untargeted metabolomics by mass spectrometry is the method of choice for unbiased analysis of molecules in complex samples of biological, clinical or environmental relevance. The exceptional versatility and sensitivity of modern high-resolution instruments allows profiling of thousands of known and unknown molecules in parallel. Inter-batch differences constitute a common and unresolved problem in untargeted metabolomics, and hinder the analysis of multi-batch studies or the intercomparison of experiments. RESULTS: We present a new method, Regularized Adversarial Learning Preserving Similarity (RALPS), for the normalization of multi-batch untargeted metabolomics data. RALPS builds on deep adversarial learning with a three-term loss function that mitigates batch effects while preserving biological identity, spectral properties and coefficients of variation. Using two large metabolomics datasets, we showcase the superior performance of RALPS as compared with six state-of-the-art methods for batch correction. Further, we demonstrate that RALPS scales well, is robust, deals with missing values and can handle different experimental designs. AVAILABILITY AND IMPLEMENTATION: https://github.com/zamboni-lab/RALPS. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

A system suitability testing platform for untargeted, high-resolution mass spectrometry.

The broad coverage of untargeted metabolomics poses fundamental challenges for the harmonization of measurements along time, even if they originate from the very same instrument. Internal isotopic standards can hardly cover the chemical complexity of study samples. Therefore, they are insufficient for normalizing data a posteriori as done for targeted metabolomics. Instead, it is crucial to verify instrument's performance a priori, that is, before samples are injected. Here, we propose a system suitability testing platform for time-of-flight mass spectrometers independent of liquid chromatography. It includes a chemically defined quality control mixture, a fast acquisition method, software for extracting ca. 3,000 numerical features from profile data, and a simple web service for monitoring. We ran a pilot for 21 months and present illustrative results for anomaly detection or learning causal relationships between the spectral features and machine settings. Beyond mere detection of anomalies, our results highlight several future applications such as 1) recommending instrument retuning strategies to achieve desired values of quality indicators, 2) driving preventive maintenance, and 3) using the obtained, detailed spectral features for posterior data harmonization.