Linear discriminant analysis

Nature Reviews Methods Primers, Год журнала: 2024, Номер 4(1)

Опубликована: Сен. 26, 2024

Язык: Английский

---

In Silico Approach for Antibacterial Discovery: PTML Modeling of Virtual Multi-Strain Inhibitors Against Staphylococcus aureus

Pharmaceuticals, Год журнала: 2025, Номер 18(2), С. 196 - 196

Опубликована: Янв. 31, 2025

Background/Objectives: Infectious diseases caused by Staphylococcus aureus (S. aureus) have become alarming health issues worldwide due to the ever-increasing emergence of multidrug resistance. In silico approaches can accelerate identification and/or design versatile antibacterial chemicals with ability target multiple S. strains varying degrees drug Here, we develop a perturbation theory machine learning model based on multilayer perceptron neural network (PTML-MLP) for prediction and virtual inhibitors against strains. Methods: To PTML-MLP model, chemical biological data associated activity were retrieved from ChEMBL database. We applied Box-Jenkins approach convert topological indices into multi-label graph-theoretical indices; latter used as inputs creation model. Results: The exhibited accuracy higher than 80% in both training test sets. physicochemical structural interpretation was performed through fragment-based (FBTD) approach. Such interpretations permitted analysis different molecular fragments favorable contributions multi-strain four new drug-like molecules using building blocks. designed predicted/confirmed our PTML diverse strains, thus representing promising chemotypes be considered future synthesis testing anti-S. agents. Conclusions: This work envisages applications modeling early discovery related antimicrobial research areas.

Язык: Английский

---

Perturbation Theory Machine Learning Model for Phenotypic Early Antineoplastic Drug Discovery: Design of Virtual Anti-Lung-Cancer Agents

Applied Sciences, Год журнала: 2024, Номер 14(20), С. 9344 - 9344

Опубликована: Окт. 14, 2024

Lung cancer is the most diagnosed malignant neoplasm worldwide and it associated with great mortality. Currently, developing antineoplastic agents a challenging, time-consuming, costly process. Computational methods can speed up early discovery of anti-lung-cancer chemicals. Here, we report perturbation theory machine learning model based on multilayer perceptron (PTML-MLP) for phenotypic drug discovery, enabling rational design prediction new molecules as virtual versatile inhibitors multiple lung cell lines. The PTML-MLP achieved an accuracy above 80%. We applied fragment-based topological (FBTD) approach to physicochemically structurally interpret model. This enabled extraction suitable fragments positive influence activity against different By following aforementioned interpretations, could assemble several four novel molecules, which were predicted by agents. Such predictions potent multi-cellular anticancer diverse lines rigorously confirmed well-established screening tool reported in literature. present work envisages opportunities application PTML models accelerate from assays.

Язык: Английский

---

Artificial Intelligence-Driven Modeling for Hydrogel Three-Dimensional Printing: Computational and Experimental Cases of Study

Polymers, Год журнала: 2025, Номер 17(1), С. 121 - 121

Опубликована: Янв. 6, 2025

Determining the values of various properties for new bio-inks 3D printing is a very important task in design materials. For this purpose, large number experimental works have been consulted, and database with more than 1200 bioprinting tests has created. These cover different combinations conditions terms print pressure, temperature, needle values, example. data are difficult to deal determining optimize analyze options. The best model demonstrated specificity (Sp) 88.4% sensitivity (Sn) 86.2% training series while achieving an Sp 85.9% Sn 80.3% external validation series. This utilizes operators based on perturbation theory complexity data. comparative purposes, neural networks used, similar results obtained. developed tool could easily be applied predict assays silico. findings significantly improve efficiency accuracy predictive models without resorting trial-and-error tests, thereby saving time funds. Ultimately, may help pave way advances personalized medicine tissue engineering.

Язык: Английский

---

Perturbation-Theory Machine Learning for Multi-Objective Antibacterial Discovery: Current Status and Future Perspectives

Applied Sciences, Год журнала: 2025, Номер 15(3), С. 1166 - 1166

Опубликована: Янв. 24, 2025

Antibacterial drugs (commonly known as antibiotics) are essential for eradicating bacterial infections. Nowadays, antibacterial discovery has become an imperative need due to the lack of efficacious antibiotics, ever-increasing development multi-drug resistance (MDR), and withdrawal many pharmaceutical industries from programs. Currently, drug is widely recognized a multi-objective optimization problem where computational approaches could play pivotal role, enabling identification novel versatile agents. Yet, tackling complex phenomena such multi-genic nature infections MDR major disadvantage most modern methods. To best our knowledge, perturbation-theory machine learning (PTML) appears be only approach capable overcoming aforementioned limitation. The present review discusses PTML modeling suitable cutting-edge in discovery. In this sense, we focus attention on application models prediction and/or design multi-target (multi-protein or multi-strain) inhibitors context small organic molecules, peptide design, metal-containing nanoparticles. Additionally, highlight future applications drug-like chemotypes with multi-protein multi-strain activity.

Язык: Английский

---

Privileged small molecules against neglected tropical diseases: A perspective from structure activity relationships

European Journal of Medicinal Chemistry, Год журнала: 2024, Номер 271, С. 116396 - 116396

Опубликована: Апрель 16, 2024

Язык: Английский

---

A Review on Revolutionizing Healthcare Technologies with AI and ML Applications in Pharmaceutical Sciences

Drugs and Drug Candidates, Год журнала: 2025, Номер 4(1), С. 9 - 9

Опубликована: Март 4, 2025

Background/Objectives: The integration of Artificial Intelligence (AI) and Machine Learning (ML) in pharmaceutical research development is transforming the industry by improving efficiency effectiveness across drug discovery, development, healthcare delivery. This review explores diverse applications AI ML, emphasizing their role predictive modeling, repurposing, lead optimization, clinical trials. Additionally, highlights AI’s contributions to regulatory compliance, pharmacovigilance, personalized medicine while addressing ethical considerations. Methods: A comprehensive literature was conducted assess impact ML various domains. Research articles, case studies, reports were analyzed examine AI-driven advancements computational chemistry, trials, safety, supply chain management. Results: have demonstrated significant research, including improved target identification, accelerated discovery through generative models, enhanced structure-based design via molecular docking QSAR modeling. In streamlines patient recruitment, predicts trial outcomes, enables real-time monitoring. maintenance, process inventory management manufacturing chains. Furthermore, has revolutionized enabling precise treatment strategies genomic data analysis, biomarker diagnostics. Conclusions: are reshaping offering innovative solutions care. enhances outcomes operational efficiencies raising challenges that require transparent, accountable applications. Future will rely on collaborative efforts ensure its responsible implementation, ultimately driving continued transformation sector.

Язык: Английский

---

Ensemble Techniques for Predictive Modeling of Leishmanial Activity via Molecular Fingerprints

Research Square (Research Square), Год журнала: 2025, Номер unknown

Опубликована: Апрель 23, 2025

Background:Leishmaniasis, a neglected tropical disease caused by Leishmania protozoan parasites and transmitted sandflies, poses significant global health challenge, especially in resource-limited environments. The life cycle of the parasite includes crucial amastigote promastigote stages, each contributing importantly to infection process. current therapies for leishmaniasis face limitations due considerable side effects rise drug-resistant strains, underscoring pressing need new, effective, safe treatment options. \textcolor{red}{Recent advancements vaccine development include live attenuated vaccines, recombinant use synthetic biology. These approaches aim induce robust immune responses while ensuring safety. Controlled human studies are also being explored accelerate development. However, licensed remains elusive.} Method:This study introduces novel method drug discovery targeting leishmaniasis, employing machine learning cheminformatics forecast efficacy compounds against promastigotes. A detailed dataset consisting 65,057 molecules sourced from PubChem database is utilized, with Alamar Blue-based assay applied assess susceptibility. data encoding relies on molecular fingerprints derived Simplified Molecular Input Line Entry System (SMILES) notations. We employed three distinct fingerprint algorithms, Avalon, MACCS Key, Pharmacophore, models. Various including random forest, multilayer perceptron, gradient boosting, decision tree, utilized create models that effectively classify as either active or inactive based their structural chemical characteristics, which could significantly impact process leishmaniasis. Results: additionally introduced model ensembles, achieving peak accuracy 83.65% an area under curve 0.8367. This offers promise enhancing efforts focused tackling issue Conclusion: Furthermore, proposed approach has potential serve framework addressing other overlooked diseases, offering promising alternative conventional methods associated difficulties.

Язык: Английский

---

Perturbation-Theory Machine Learning for Multi-Target Drug Discovery in Modern Anticancer Research

Current Issues in Molecular Biology, Год журнала: 2025, Номер 47(5), С. 301 - 301

Опубликована: Апрель 25, 2025

Cancers constitute a group of biological complex diseases, which are associated with great prevalence and mortality. These medical conditions very difficult to tackle due their multi-factorial nature, includes ability evade the immune system become resistant current anticancer agents. There is pressing need search for novel agents multi-target modes action and/or multi-cell inhibition versatility, can translate into more efficacious safer chemotherapeutic treatments. Computational methods paramount importance accelerate drug discovery in cancer research but most them have several disadvantages such as use limited structural information through homogeneous datasets chemicals, prediction activity against single target, lack interpretability. This mini-review discusses emergence, development, application perturbation-theory machine learning (PTML) cutting-edge approach capable overcoming aforementioned limitations context small molecule discovery. Here, we analyze promising investigations on PTML modeling spanning over decade enable versatile We highlight potential while envisaging future applications modeling.

Язык: Английский

---

Unraveling the artificial intelligence role in drug discovery and pharmaceutical product design: an opportunity and challenges

Discover Artificial Intelligence, Год журнала: 2025, Номер 5(1)

Опубликована: Май 25, 2025

Язык: Английский

---