Pub Date : 2026-01-18DOI: 10.1016/j.ailsci.2026.100156
Ivon Acosta-Ramirez , Ferhat Sadak , Sruti Das Choudhury , James Thomson , Salome Perez-Rosero , Portia N.A. Plange , Sofia E. Morales-Mendivelso , Nicole M. Iverson
Detecting the spatial release of extracellular nitric oxide (NO) is essential for understanding the dynamics in cell communication for physiological and pathological processes. This study presents an innovative methodology that integrates fluorescence-based sensing platforms utilizing single walled carbon nanotubes (SWNT) with machine learning models to expedite the spatial data analysis of extracellular analytes. The deep learning model You Only Look Once (YOLOv8) segmentation achieves accurate cell identification across diverse morphologies and clustered cell groups, with a recall of 98% and a precision of 83%. The spatial analysis of extracellular NO is achieved by extracting the cell contour coordinates from the YOLO-identified cells and translocating the boundaries onto SWNT fluorescence files. The model enables rapid analysis for multiple cells across numerous images, with 100 image pairs completed in just 68 s. The combination of nanotechnology with automated neural network-based cell detection establishes a robust sensing framework with pixel-level spatial resolution of NO dynamics, delivering critical insights into cellular communication and holding promising implications for diagnostic and therapeutic applications.
检测细胞外一氧化氮（NO）的空间释放对于理解生理和病理过程中细胞通讯的动力学至关重要。本研究提出了一种创新的方法，将利用单壁碳纳米管（SWNT）的基于荧光的传感平台与机器学习模型相结合，以加快细胞外分析物的空间数据分析。深度学习模型You Only Look Once （YOLOv8）分割在不同形态和集群细胞群中实现了准确的细胞识别，召回率为98%，精度为83%。细胞外NO的空间分析是通过从yolo识别的细胞中提取细胞轮廓坐标并将边界转移到SWNT荧光文件中来实现的。该模型可以快速分析众多图像中的多个细胞，只需68秒即可完成100对图像。纳米技术与基于自动神经网络的细胞检测相结合，建立了一个具有NO动态像素级空间分辨率的强大传感框架，为细胞通信提供了关键见解，并为诊断和治疗应用带来了希望。
{"title":"Development of a deep neural network model for simultaneous analysis of extracellular analyte gradients for a population of cells","authors":"Ivon Acosta-Ramirez , Ferhat Sadak , Sruti Das Choudhury , James Thomson , Salome Perez-Rosero , Portia N.A. Plange , Sofia E. Morales-Mendivelso , Nicole M. Iverson","doi":"10.1016/j.ailsci.2026.100156","DOIUrl":"10.1016/j.ailsci.2026.100156","url":null,"abstract":"<div><div>Detecting the spatial release of extracellular nitric oxide (NO) is essential for understanding the dynamics in cell communication for physiological and pathological processes. This study presents an innovative methodology that integrates fluorescence-based sensing platforms utilizing single walled carbon nanotubes (SWNT) with machine learning models to expedite the spatial data analysis of extracellular analytes. The deep learning model You Only Look Once (YOLOv8) segmentation achieves accurate cell identification across diverse morphologies and clustered cell groups, with a recall of 98% and a precision of 83%. The spatial analysis of extracellular NO is achieved by extracting the cell contour coordinates from the YOLO-identified cells and translocating the boundaries onto SWNT fluorescence files. The model enables rapid analysis for multiple cells across numerous images, with 100 image pairs completed in just 68 s. The combination of nanotechnology with automated neural network-based cell detection establishes a robust sensing framework with pixel-level spatial resolution of NO dynamics, delivering critical insights into cellular communication and holding promising implications for diagnostic and therapeutic applications.</div></div>","PeriodicalId":72304,"journal":{"name":"Artificial intelligence in the life sciences","volume":"9 ","pages":"Article 100156"},"PeriodicalIF":5.4,"publicationDate":"2026-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146038571","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-09DOI: 10.1016/j.ailsci.2026.100154
Atsushi Yoshimori , Jürgen Bajorath
Compound optimization is of central relevance in medicinal chemistry. We introduce a new machine learning framework for iterative chemical optimization that integrates compound potency predictions, the explanation of predictions, and generative modeling and that is applicable to individual compounds. The approach identifies substituents in active compounds that limit their potency and iteratively replaces these substituents with others supporting potency increases. In proof-of-concept calculations, the methodology effectively optimizes compound potency. Furthermore, the optimization framework is combined with a large language model via the model concept protocol to generate an AI agent system for interactive optimization. The system is shown to successfully carry out optimization tasks of increasing complexity based on simple prompts, without the need for additional fine-tuning. The interactive computational optimization approach is accessible to non-experts and expected to be of particular interest for practical medicinal chemistry.
{"title":"A client-enhanced language model for interactive compound optimization guided by explainable artificial intelligence","authors":"Atsushi Yoshimori , Jürgen Bajorath","doi":"10.1016/j.ailsci.2026.100154","DOIUrl":"10.1016/j.ailsci.2026.100154","url":null,"abstract":"<div><div>Compound optimization is of central relevance in medicinal chemistry. We introduce a new machine learning framework for iterative chemical optimization that integrates compound potency predictions, the explanation of predictions, and generative modeling and that is applicable to individual compounds. The approach identifies substituents in active compounds that limit their potency and iteratively replaces these substituents with others supporting potency increases. In proof-of-concept calculations, the methodology effectively optimizes compound potency. Furthermore, the optimization framework is combined with a large language model via the model concept protocol to generate an AI agent system for interactive optimization. The system is shown to successfully carry out optimization tasks of increasing complexity based on simple prompts, without the need for additional fine-tuning. The interactive computational optimization approach is accessible to non-experts and expected to be of particular interest for practical medicinal chemistry.</div></div>","PeriodicalId":72304,"journal":{"name":"Artificial intelligence in the life sciences","volume":"9 ","pages":"Article 100154"},"PeriodicalIF":5.4,"publicationDate":"2026-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145980264","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-03DOI: 10.1016/j.ailsci.2026.100153
Ondrej Krejcar , Jamaluddin Abdullah , Hamidreza Namazi
The growing adoption of artificial intelligence (AI) in life sciences has been paralleled by growing concerns regarding transparency, interpretability, and trustworthiness of predictive models. While explainable artificial intelligence (XAI) has emerged as a powerful framework to bridge this gap, its practical deployment continues to face substantial technical, ethical, and regulatory barriers. This review provides a comprehensive overview of the challenges associated with implementing XAI in life science applications—including data complexity, model heterogeneity, computational costs, clinical integration, and ethical considerations—and discusses potential solutions and strategies to address them. By mapping recent advances in methodological approaches, regulatory frameworks, and interdisciplinary collaborations, we highlight a roadmap for embedding explainability into the AI lifecycle. The paper concludes with future perspectives on harmonizing interpretability with predictive performance in critical domains such as drug discovery, medical diagnostics, and bioinformatics.
{"title":"Implementing XAI in life sciences: Key challenges and pathways to solutions","authors":"Ondrej Krejcar , Jamaluddin Abdullah , Hamidreza Namazi","doi":"10.1016/j.ailsci.2026.100153","DOIUrl":"10.1016/j.ailsci.2026.100153","url":null,"abstract":"<div><div>The growing adoption of artificial intelligence (AI) in life sciences has been paralleled by growing concerns regarding transparency, interpretability, and trustworthiness of predictive models. While explainable artificial intelligence (XAI) has emerged as a powerful framework to bridge this gap, its practical deployment continues to face substantial technical, ethical, and regulatory barriers. This review provides a comprehensive overview of the challenges associated with implementing XAI in life science applications—including data complexity, model heterogeneity, computational costs, clinical integration, and ethical considerations—and discusses potential solutions and strategies to address them. By mapping recent advances in methodological approaches, regulatory frameworks, and interdisciplinary collaborations, we highlight a roadmap for embedding explainability into the AI lifecycle. The paper concludes with future perspectives on harmonizing interpretability with predictive performance in critical domains such as drug discovery, medical diagnostics, and bioinformatics.</div></div>","PeriodicalId":72304,"journal":{"name":"Artificial intelligence in the life sciences","volume":"9 ","pages":"Article 100153"},"PeriodicalIF":5.4,"publicationDate":"2026-01-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145980263","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-30DOI: 10.1016/j.ailsci.2025.100152
Edvin Forsgren , Jonne Rietdijk , David Holmberg , Julia Juneblad , Bianca Migliori , Martin M. Johansson , Jordi Carreras-Puigvert , Johan Trygg , Gillian Lovell , Ola Spjuth , Pär Jonsson
Morphological profiling is a common approach to investigate the modes of action (MOAs) of compounds. Most methods rely on fixed-cell assays, which provide only a single snapshot at a predefined time point and overlook the dynamic nature of cellular responses. In contrast, live-cell imaging tracks responses over time, offering deeper insight into compound-specific effects and mechanisms; however, time-series analysis of image data remains challenging due to limited analytical tools.
We present Live Cell Temporal Profiling (LCTP), a workflow for morphological profiling of label-free live-cell time series data that yields interpretable, biologically relevant results. We showcase LCTP in an MOA classification study using label-free data. The workflow integrates established deep-learning components, cell segmentation, live/dead classification, and single-cell feature extraction, with data-driven models to capture MOA-specific temporal phenotypes and produce time-resolved profiles that can be compared across compounds and cell lines.
We assess MOA classification performance using double-blinded cross-validation simulating a real-world screening scenario. LCTP significantly improves MOA classification over single–time point analysis, consistently across both cell lines used in the study. Time-resolved phenotypic modelling reveals transient, sustained, and delayed responses, clarifying compound-specific temporal effects and mechanisms across MOAs.
The presented workflow is modular: each step removes irrelevant information, enriching signal, and enabling straightforward updates as technologies evolve and as new technologies become available, while supporting reuse across studies broadly. We believe LCTP adds substantial value to high-throughput compound screening, showing that live-cell imaging combined with this workflow yields informative visualizations of temporal effects and improved MOA classification.
{"title":"The time dimension matters: Improving mode of action classification with live-cell imaging","authors":"Edvin Forsgren , Jonne Rietdijk , David Holmberg , Julia Juneblad , Bianca Migliori , Martin M. Johansson , Jordi Carreras-Puigvert , Johan Trygg , Gillian Lovell , Ola Spjuth , Pär Jonsson","doi":"10.1016/j.ailsci.2025.100152","DOIUrl":"10.1016/j.ailsci.2025.100152","url":null,"abstract":"<div><div>Morphological profiling is a common approach to investigate the modes of action (MOAs) of compounds. Most methods rely on fixed-cell assays, which provide only a single snapshot at a predefined time point and overlook the dynamic nature of cellular responses. In contrast, live-cell imaging tracks responses over time, offering deeper insight into compound-specific effects and mechanisms; however, time-series analysis of image data remains challenging due to limited analytical tools.</div><div>We present Live Cell Temporal Profiling (LCTP), a workflow for morphological profiling of label-free live-cell time series data that yields interpretable, biologically relevant results. We showcase LCTP in an MOA classification study using label-free data. The workflow integrates established deep-learning components, cell segmentation, live/dead classification, and single-cell feature extraction, with data-driven models to capture MOA-specific temporal phenotypes and produce time-resolved profiles that can be compared across compounds and cell lines.</div><div>We assess MOA classification performance using double-blinded cross-validation simulating a real-world screening scenario. LCTP significantly improves MOA classification over single–time point analysis, consistently across both cell lines used in the study. Time-resolved phenotypic modelling reveals transient, sustained, and delayed responses, clarifying compound-specific temporal effects and mechanisms across MOAs.</div><div>The presented workflow is modular: each step removes irrelevant information, enriching signal, and enabling straightforward updates as technologies evolve and as new technologies become available, while supporting reuse across studies broadly. We believe LCTP adds substantial value to high-throughput compound screening, showing that live-cell imaging combined with this workflow yields informative visualizations of temporal effects and improved MOA classification.</div></div>","PeriodicalId":72304,"journal":{"name":"Artificial intelligence in the life sciences","volume":"9 ","pages":"Article 100152"},"PeriodicalIF":5.4,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145980265","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Drug discovery is a complex, resource-intensive process requiring significant time and cost to bring new medicines to patients. Many generative models aim to accelerate drug discovery, but few produce synthetically accessible molecules. Conversely, synthesis-focused models do not leverage the 3D information crucial for effective drug design. We introduce SynthFormer, a novel machine learning model that generates fully synthesizable molecules, structured as synthetic trees, by introducing both 3D information and pharmacophores as input. SynthFormer features a 3D equivariant graph neural network to encode pharmacophores, followed by a Transformer-based synthesis-aware decoding mechanism for constructing synthetic trees as a sequence of tokens. This provides capabilities for designing active molecules based on pharmacophores, exploring the local synthesizable chemical space around hit molecules and optimizing their properties. We demonstrate its effectiveness through various challenging tasks, including designing active compounds for a range of proteins, performing hit expansion and optimizing molecular properties.
{"title":"SynthFormer: Equivariant pharmacophore-based generation of synthesizable molecules for ligand-based drug design","authors":"Zygimantas Jocys , Zhanxing Zhu , Henriette M.G. Willems , Katayoun Farrahi","doi":"10.1016/j.ailsci.2025.100148","DOIUrl":"10.1016/j.ailsci.2025.100148","url":null,"abstract":"<div><div>Drug discovery is a complex, resource-intensive process requiring significant time and cost to bring new medicines to patients. Many generative models aim to accelerate drug discovery, but few produce synthetically accessible molecules. Conversely, synthesis-focused models do not leverage the 3D information crucial for effective drug design. We introduce SynthFormer, a novel machine learning model that generates fully synthesizable molecules, structured as synthetic trees, by introducing both 3D information and pharmacophores as input. SynthFormer features a 3D equivariant graph neural network to encode pharmacophores, followed by a Transformer-based synthesis-aware decoding mechanism for constructing synthetic trees as a sequence of tokens. This provides capabilities for designing active molecules based on pharmacophores, exploring the local synthesizable chemical space around hit molecules and optimizing their properties. We demonstrate its effectiveness through various challenging tasks, including designing active compounds for a range of proteins, performing hit expansion and optimizing molecular properties.</div></div>","PeriodicalId":72304,"journal":{"name":"Artificial intelligence in the life sciences","volume":"9 ","pages":"Article 100148"},"PeriodicalIF":5.4,"publicationDate":"2025-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145898134","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01DOI: 10.1016/j.ailsci.2025.100147
Yinli Shi , Jun Liu , Sicun Wang , Shuang Guan , Muzhi Li , Yanan Yu , Hu Yang , Wei Yang , Bing Li , Weibin Yang , Xuezhong Zhou , Zhong Wang
Although combination drug therapies hold great promise for complex diseases, their development is hindered by the complexity of biological networks and the combinatorial explosion of possible drug interactions. Deep learning (DL) models offer a transformative solution by integrating multimodal data and biomedical networks to predict drug combination synergy with high accuracy. These models automatically extract complex patterns from high-dimensional data, overcoming limitations of conventional methods, accelerating rational combination discovery. Here, we systematically examined diverse network-based DL frameworks, analyzing how increasing structural complexity enhances prediction performance while maintaining interpretability. While current methodologies show encouraging results, challenges remain in data quality, model generalization, and clinical translation. Here, we highlight pivotal studies demonstrating in different DL models’ potential, outlines their key limitations, and discusses future directions including multimodal learning and mechanistic interpretability, to establish multilayer DL model as a cornerstone of next-generation drug combination discovery.
{"title":"Drug discovery of synergistic combinations via multilayer deep learning models:Advances and challenges","authors":"Yinli Shi , Jun Liu , Sicun Wang , Shuang Guan , Muzhi Li , Yanan Yu , Hu Yang , Wei Yang , Bing Li , Weibin Yang , Xuezhong Zhou , Zhong Wang","doi":"10.1016/j.ailsci.2025.100147","DOIUrl":"10.1016/j.ailsci.2025.100147","url":null,"abstract":"<div><div>Although combination drug therapies hold great promise for complex diseases, their development is hindered by the complexity of biological networks and the combinatorial explosion of possible drug interactions. Deep learning (DL) models offer a transformative solution by integrating multimodal data and biomedical networks to predict drug combination synergy with high accuracy. These models automatically extract complex patterns from high-dimensional data, overcoming limitations of conventional methods, accelerating rational combination discovery. Here, we systematically examined diverse network-based DL frameworks, analyzing how increasing structural complexity enhances prediction performance while maintaining interpretability. While current methodologies show encouraging results, challenges remain in data quality, model generalization, and clinical translation. Here, we highlight pivotal studies demonstrating in different DL models’ potential, outlines their key limitations, and discusses future directions including multimodal learning and mechanistic interpretability, to establish multilayer DL model as a cornerstone of next-generation drug combination discovery.</div></div>","PeriodicalId":72304,"journal":{"name":"Artificial intelligence in the life sciences","volume":"8 ","pages":"Article 100147"},"PeriodicalIF":5.4,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145623310","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01DOI: 10.1016/j.ailsci.2025.100146
Jason Granstedt, Prabhat Kc, Rucha Deshpande, Victor Garcia, Aldo Badano
{"title":"Corrigendum to “Hallucinations in medical devices” [Artif. Intell. Life Sci. 8 (2025) 100145]","authors":"Jason Granstedt, Prabhat Kc, Rucha Deshpande, Victor Garcia, Aldo Badano","doi":"10.1016/j.ailsci.2025.100146","DOIUrl":"10.1016/j.ailsci.2025.100146","url":null,"abstract":"","PeriodicalId":72304,"journal":{"name":"Artificial intelligence in the life sciences","volume":"8 ","pages":"Article 100146"},"PeriodicalIF":5.4,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145747425","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01DOI: 10.1016/j.ailsci.2025.100149
Marina Bilotta , Roberta Rocca , Stefano Alcaro
The integration of artificial intelligence (AI) into the drug discovery pipeline is redefining pharmaceutical research by enhancing efficiency, predictive accuracy, and innovation. Traditional drug development, constrained by high costs, long timelines, and low success rates, is being transformed through deep learning, predictive modeling, and explainable AI (XAI). These tools accelerate target identification, lead optimization, and drug repurposing by enabling high-throughput interpretation of multi-omics datasets spanning genomics, proteomics, and metabolomics. Generative models, including variational autoencoders (VAEs), generative adversarial networks (GANs), and transformer-based architectures, enable the de novo design of bioactive compounds, while reinforcement learning refines molecular properties. Structure-based drug design has been advanced by graph neural networks (GNNs) and convolutional neural networks (CNNs), improving virtual screening and binding affinity prediction. The coupling of AI with quantum chemistry enhances molecular property estimation, reducing reliance on experimental validation. AI-driven prediction of drug–target interactions (DTIs) supports both repurposing efforts and pharmacovigilance. This review presents a polypharmacology-aware, feedback-to-discovery framework, in which translational signals, such as biomarkers, molecular subtypes, and pathway constraints, are reintegrated into target selection and compound optimization to enhance decision quality. Unlike previous reviews focused on isolated AI applications, it offers a unified, end-to-end synthesis spanning target discovery to regulatory translation. We distinguish foundation models that learn transferable molecular representations from generative models that synthesize new compounds. Together with multimodal learning, explainable AI, and closed-loop design–make–test–learn systems linking molecular design to automated synthesis, these advances outline a mechanism-informed roadmap for AI-driven discovery across the modern pharmaceutical pipeline.
{"title":"Next-generation drug discovery: The AI revolution in pharmaceutical research","authors":"Marina Bilotta , Roberta Rocca , Stefano Alcaro","doi":"10.1016/j.ailsci.2025.100149","DOIUrl":"10.1016/j.ailsci.2025.100149","url":null,"abstract":"<div><div>The integration of artificial intelligence (AI) into the drug discovery pipeline is redefining pharmaceutical research by enhancing efficiency, predictive accuracy, and innovation. Traditional drug development, constrained by high costs, long timelines, and low success rates, is being transformed through deep learning, predictive modeling, and explainable AI (XAI). These tools accelerate target identification, lead optimization, and drug repurposing by enabling high-throughput interpretation of multi-omics datasets spanning genomics, proteomics, and metabolomics. Generative models, including variational autoencoders (VAEs), generative adversarial networks (GANs), and transformer-based architectures, enable the de novo design of bioactive compounds, while reinforcement learning refines molecular properties. Structure-based drug design has been advanced by graph neural networks (GNNs) and convolutional neural networks (CNNs), improving virtual screening and binding affinity prediction. The coupling of AI with quantum chemistry enhances molecular property estimation, reducing reliance on experimental validation. AI-driven prediction of drug–target interactions (DTIs) supports both repurposing efforts and pharmacovigilance. This review presents a polypharmacology-aware, feedback-to-discovery framework, in which translational signals, such as biomarkers, molecular subtypes, and pathway constraints, are reintegrated into target selection and compound optimization to enhance decision quality. Unlike previous reviews focused on isolated AI applications, it offers a unified, end-to-end synthesis spanning target discovery to regulatory translation. We distinguish foundation models that learn transferable molecular representations from generative models that synthesize new compounds. Together with multimodal learning, explainable AI, and closed-loop design–make–test–learn systems linking molecular design to automated synthesis, these advances outline a mechanism-informed roadmap for AI-driven discovery across the modern pharmaceutical pipeline.</div></div>","PeriodicalId":72304,"journal":{"name":"Artificial intelligence in the life sciences","volume":"8 ","pages":"Article 100149"},"PeriodicalIF":5.4,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145693436","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-24DOI: 10.1016/j.ailsci.2025.100145
Jason Granstedt, Prabhat Kc, Rucha Deshpande, Victor Garcia, Aldo Badano
Computer methods in medical devices are frequently imperfect and are known to produce errors in clinical or diagnostic tasks. However, when deep learning and data-based approaches yield output that exhibit errors, the devices are frequently said to hallucinate. Drawing from theoretical developments and empirical studies in multiple medical device areas, we introduce a practical and universal definition that denotes hallucinations as a type of error that is plausible and can be either impactful or benign to the task at hand. The definition aims at facilitating the evaluation of medical devices that suffer from hallucinations across product areas. Using examples from imaging and non-imaging applications, we explore how the proposed definition relates to evaluation methodologies and discuss existing approaches for minimizing the prevalence of hallucinations.
{"title":"Hallucinations in medical devices","authors":"Jason Granstedt, Prabhat Kc, Rucha Deshpande, Victor Garcia, Aldo Badano","doi":"10.1016/j.ailsci.2025.100145","DOIUrl":"10.1016/j.ailsci.2025.100145","url":null,"abstract":"<div><div>Computer methods in medical devices are frequently imperfect and are known to produce errors in clinical or diagnostic tasks. However, when deep learning and data-based approaches yield output that exhibit errors, the devices are frequently said to hallucinate. Drawing from theoretical developments and empirical studies in multiple medical device areas, we introduce a practical and universal definition that denotes hallucinations as a type of error that is plausible and can be either impactful or benign to the task at hand. The definition aims at facilitating the evaluation of medical devices that suffer from hallucinations across product areas. Using examples from imaging and non-imaging applications, we explore how the proposed definition relates to evaluation methodologies and discuss existing approaches for minimizing the prevalence of hallucinations.</div></div>","PeriodicalId":72304,"journal":{"name":"Artificial intelligence in the life sciences","volume":"8 ","pages":"Article 100145"},"PeriodicalIF":5.4,"publicationDate":"2025-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145473665","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-23DOI: 10.1016/j.ailsci.2025.100144
Serkan Eti , Serhat Yüksel , Seçil Topaloğlu Eti , Hasan Dinçer , Ozan Emre Eyupoglu
The rapid escalation of antibiotic resistance is diminishing the effectiveness of current treatments and poses a severe threat to global health security. Addressing this challenge requires identifying the most critical criteria in the antibiotic development process and determining which approaches yield the most effective results. However, the literature reveals a significant gap: few studies systematically analyze the factors that shape the effectiveness of antibiotic development, and even fewer comparatively evaluate the most efficient development strategies. This study aims to fill this gap by providing a scientific roadmap for decision-makers through the integration of artificial intelligence (AI) methods into a fuzzy multi-criteria decision-making (MCDM) framework. A total of 15 evaluation criteria and eight antibiotic development approaches were identified through a comprehensive literature review. Expert opinions were collected from five specialists in the field, and their relative importance was objectively quantified using a dimensionality reduction technique, a machine learning–based AI approach. Subsequently, criteria weights were calculated via the LOPCOW method, while antibiotic development strategies were ranked using the CODAS method. To further enhance the robustness of decision-making under uncertainty, the newly introduced Koch Snowflake fuzzy sets were integrated into the AI-driven framework, marking an additional innovation in fuzzy set theory. This hybrid model contributes to the literature by (i) enabling a holistic analysis of critical factors and effective strategies in antibiotic development, (ii) demonstrating how AI-based dimensionality reduction can be combined with fuzzy decision-making tools for more objective and precise outcomes, and (iii) offering a more comprehensive evaluation than previous studies by incorporating an extended set of criteria. The study’s findings reveal that the most important factor in the antibiotic development process is smart biosafety and computerized control systems (0.0904), while the optimal development strategy is artificial intelligence-assisted molecule discovery (0.504). Additionally, antibiotic repositioning was found to play a significant supporting role. By highlighting the value of integrating machine learning techniques and fuzzy AI frameworks into drug discovery processes, this research not only addresses a pressing issue in global health but also demonstrates the transformative potential of artificial intelligence in advancing life sciences and accelerating antibiotic innovation.
{"title":"Leveraging artificial intelligence and koch snowflake fuzzy sets to optimize antibiotic development pathways","authors":"Serkan Eti , Serhat Yüksel , Seçil Topaloğlu Eti , Hasan Dinçer , Ozan Emre Eyupoglu","doi":"10.1016/j.ailsci.2025.100144","DOIUrl":"10.1016/j.ailsci.2025.100144","url":null,"abstract":"<div><div>The rapid escalation of antibiotic resistance is diminishing the effectiveness of current treatments and poses a severe threat to global health security. Addressing this challenge requires identifying the most critical criteria in the antibiotic development process and determining which approaches yield the most effective results. However, the literature reveals a significant gap: few studies systematically analyze the factors that shape the effectiveness of antibiotic development, and even fewer comparatively evaluate the most efficient development strategies. This study aims to fill this gap by providing a scientific roadmap for decision-makers through the integration of artificial intelligence (AI) methods into a fuzzy multi-criteria decision-making (MCDM) framework. A total of 15 evaluation criteria and eight antibiotic development approaches were identified through a comprehensive literature review. Expert opinions were collected from five specialists in the field, and their relative importance was objectively quantified using a dimensionality reduction technique, a machine learning–based AI approach. Subsequently, criteria weights were calculated via the LOPCOW method, while antibiotic development strategies were ranked using the CODAS method. To further enhance the robustness of decision-making under uncertainty, the newly introduced Koch Snowflake fuzzy sets were integrated into the AI-driven framework, marking an additional innovation in fuzzy set theory. This hybrid model contributes to the literature by (i) enabling a holistic analysis of critical factors and effective strategies in antibiotic development, (ii) demonstrating how AI-based dimensionality reduction can be combined with fuzzy decision-making tools for more objective and precise outcomes, and (iii) offering a more comprehensive evaluation than previous studies by incorporating an extended set of criteria. The study’s findings reveal that the most important factor in the antibiotic development process is smart biosafety and computerized control systems (0.0904), while the optimal development strategy is artificial intelligence-assisted molecule discovery (0.504). Additionally, antibiotic repositioning was found to play a significant supporting role. By highlighting the value of integrating machine learning techniques and fuzzy AI frameworks into drug discovery processes, this research not only addresses a pressing issue in global health but also demonstrates the transformative potential of artificial intelligence in advancing life sciences and accelerating antibiotic innovation.</div></div>","PeriodicalId":72304,"journal":{"name":"Artificial intelligence in the life sciences","volume":"8 ","pages":"Article 100144"},"PeriodicalIF":5.4,"publicationDate":"2025-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145362197","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号