Unsupervised Learning in Precision Medicine: Unlocking Personalized Healthcare through AI

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(20), P. 9305 - 9305

Published: Oct. 12, 2024

Integrating Artificial Intelligence (AI) into Precision Medicine (PM) is redefining healthcare, enabling personalized treatments tailored to individual patients based on their genetic code, environment, and lifestyle. AI’s ability analyze vast complex datasets, including genomics medical records, facilitates the identification of hidden patterns correlations, which are critical for developing treatment plans. Unsupervised Learning (UL) particularly valuable in PM as it can unstructured unlabeled data uncover novel disease subtypes, biomarkers, patient stratifications. By revealing that not explicitly labeled, unsupervised algorithms enable discovery new insights mechanisms variability, advancing our understanding responses treatment. However, integration AI presents some challenges, concerns about privacy rigorous validation models clinical practice. Despite these holds immense potential revolutionize PM, offering a more personalized, efficient, effective approach healthcare. Collaboration among developers clinicians essential fully realize this ensure ethical reliable implementation This review will explore latest emerging UL technologies biomedical field with particular focus applications impact human health well-being.

Language: Английский

---

SHASI-ML: a machine learning-based approach for immunogenicity prediction in Salmonella vaccine development

Frontiers in Cellular and Infection Microbiology, Journal Year: 2025, Volume and Issue: 15

Published: Feb. 11, 2025

Introduction Accurate prediction of immunogenic proteins is crucial for vaccine development and understanding host-pathogen interactions in bacterial diseases, particularly Salmonella infections which remain a significant global health challenge. Methods We developed SHASI-ML, machine learning-based framework predicting species. The model was trained validated using curated dataset experimentally verified non-immunogenic proteins. Three distinct feature groups were extracted from protein sequences: properties, sequence-derived features, structural information. Extreme Gradient Boosting (XGBoost) algorithm employed optimization. Results SHASI-ML demonstrated robust performance identifying immunogens, achieving 89.3% precision 91.2% specificity. When applied to the enterica serovar Typhimurium proteome, identified 292 novel candidates. Global properties emerged as most influential group accuracy, followed by sequence showed superior recall F1-scores compared existing computational approaches. Discussion These findings establish an efficient tool prioritizing candidates development. By streamlining identification early process, this approach significantly reduces experimental burden associated costs. methodology can be guide optimize both research industrial-scale production vaccines, potentially accelerating more effective immunization strategies.

Language: Английский

---

Predicting therapy dropout in chronic pain management: a machine learning approach to cannabis treatment

Frontiers in Artificial Intelligence, Journal Year: 2025, Volume and Issue: 8

Published: Feb. 20, 2025

Chronic pain affects approximately 30% of the global population, posing a significant public health challenge. Despite their widespread use, traditional pharmacological treatments, such as opioids and NSAIDs, often fail to deliver adequate, long-term relief while exposing patients risks addiction adverse side effects. Given these limitations, medical cannabis has emerged promising therapeutic alternative with both analgesic anti-inflammatory properties. However, its clinical efficacy is hindered by high interindividual variability in treatment response elevated dropout rates. A comprehensive dataset integrating genetic, clinical, information was compiled from 542 Caucasian undergoing cannabis-based for chronic pain. machine learning (ML) model developed validated predict therapy dropout. To identify most influential factors driving dropout, SHapley Additive exPlanations (SHAP) analysis performed. The random forest classifier demonstrated robust performance, achieving mean accuracy 80% maximum 86%, an AUC 0.86. SHAP revealed that final VAS scores THC dosages were predictors strongly correlated increased likelihood discontinuation. In contrast, baseline benefits, CBD dosages, CC genotype rs1049353 polymorphism CNR1 gene associated improved adherence. Our findings highlight potential ML pharmacogenetics personalize therapies, improving adherence enabling more precise management This research paves way development tailored strategies maximize benefits minimizing

Language: Английский

---

Moving towards the use of artificial intelligence in pain management

European Journal of Pain, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 10, 2024

While the development of artificial intelligence (AI) technologies in medicine has been significant, their application to acute and chronic pain management not well characterized. This systematic review aims provide an overview current state AI management.

Language: Английский

---

Exploring the Potential of Nonpsychoactive Cannabinoids in the Development of Materials for Biomedical and Sports Applications

ACS Applied Bio Materials, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 20, 2024

This Perspective explores the potential of nonpsychoactive cannabinoids (NPCs) such as CBD, CBG, CBC, and CBN in developing innovative biomaterials for biomedical sports applications. It examines their physicochemical properties, anti-inflammatory, analgesic, neuroprotective effects, integration into various hydrogels, sponges, films, scaffolds. also discusses current challenges standardizing formulations, understanding long-term intrinsical regulatory landscapes. Further, it promising applications NPC-loaded materials bone regeneration, wound management, drug delivery systems, emphasizing improved biocompatibility, mechanical therapeutic efficacy demonstrated

Language: Английский

---

Three-Dimensional Quantitative Structure–Activity Relationship Study of Transient Receptor Potential Vanilloid 1 Channel Antagonists Reveals Potential for Drug Design Purposes

International Journal of Molecular Sciences, Journal Year: 2024, Volume and Issue: 25(14), P. 7951 - 7951

Published: July 21, 2024

Transient receptor potential vanilloid 1 (TRPV1) was reported to be a putative target for recovery from chronic pain, producing analgesic effects after its inhibition. A series of drug candidates were previously developed, without the ability ameliorate therapeutic outcome. Starting designed compounds, derived hybridization antagonist SB-705498 and partial agonist MDR-652, we performed virtual screening on pharmacophore model built by exploiting Cryo-EM 3D structure nanomolar in complex with human TRPV1 channel. The described three pharmacophoric features, taking advantage both bioactive pose exclusion spheres. results implemented inside 3D-QSAR model, correlating negative decadic logarithm inhibition rate ligands. After validation obtained new compounds introducing key modifications original scaffold. Again, determined compounds' binding poses alignment predicted their rates validated model. values resulted being even more promising than parent demonstrating that ongoing research still leaves much room improvement.

Language: Английский

---