搜索结果：Clinical and translational medicine

Spatiotemporal molecular medicine is a new discipline of medicine to precisely understand the pathogenesis, history, and epidemiology, and aims to achieve prevention, diagnosis, and therapy at molecular levels for human diseases in multidimensional aspects related to a spatial and temporal context. The human body is a three-dimensional object with the space of length, width, and height, while the spatiotemporal concept demonstrates a four-dimensional continuum with the concept of time. Clinical spatialization of a disease should be considered at the levels of genetics (e.g., family tree, pedigree, germline heterogeneity), populations (e.g., regional distribution, epidemiological map), and individuals (e.g., inter- and intraorgan location and variation). Clinical temporalization covers the historical process and progression of disease occurrence and development, duration-based alteration of clinical phenomes, and dynamics of patient responses to therapy. Clinical trans-omics was proposed as a new approach to understand the disease from the cross-points among different layers of networks generated from the integration of clinical phenomics with molecular multi-omics.1 Clinical trans-omics provide three-dimensional information on clinical phenome(s) corresponding to molecular elements or the reverse, the understanding of molecule-based dimension in correlation with phenomes, and the potential of disease phenome-specific diagnostic biomarkers and therapeutic targets. In comparison, spatiotemporal molecular medicine presents a four-dimensional and dynamical picture of the disease by integrating clinical spatialization, temporalization, phenome, and molecular multi-omics for disease diagnosis, therapy, and prognosis. We propose that spatiotemporal molecular medicine will be an independent and merging discipline to meet the rapid development of medicine. The present editorial especially highlights spatiotemporal molecular medicine as a new vision and platform to understand the importance of spatial and temporal transcriptomics/proteomics, positioning of cell–cell interaction and communication, organogenesis, and organ development for diagnosis and therapy. We also address the difference between spatiotemporal medicine and spatiotemporal molecular medicine, potential availability of spatiotemporal molecular omics measurements and analyses, and need of artificial intelligence and computerized models. Spatiotemporal molecular omics are a major part of spatiotemporal molecular medicine. For example, transcriptomic profiles of tissues, isolated single category of cells, or single cells demonstrate tissue- or cell-specific transcriptional function as one-dimensional information of mRNA expression. Ståhl et al. labeled arrayed reverse transcription primers with unique barcodes, in situ hybridized positioning mRNA on tissue sections with those primers, as well as visually and quantitively imaged and analyzed transcriptomic profiles on tissue surface, coined as spatial transcriptomics, in order to define the exact position of transcriptional alterations with the tissue and cells.2 This was a revolutionary development to clearly define the exact locations of cells with up-/downregulated RNA expression and provide two-dimensional information. Furthermore, analyses of single cell and spatial transcriptomics were developed by integration of isolated single-cell RNA sequencing and spatial transcriptomics as a new approach to reconstruct tissue visualization and neighbor structure and define cell–cell interactions in tissues, spatial relations between cell locations and transcriptional phenomes, and evolutions of lineages and populations. Single cell and spatial transcriptomics were proposed to have special values to deeply understand mechanisms of cell function and pathogenesis of the disease and precisely identify spatial responses of target cells to therapy.3 In addition, spatiotemporal proteomics provide an in-depth cell-type- and region-specific proteomic catalog of the tissue section using the peptide identifications from the soluble or insoluble proteomes by single-run liquid chromatography–tandem mass spectrometry. Spatiotemporal proteomics demonstrated that three selected clusters showed distinct protein expression across four regions with functional brain region specificity, responsible for the progression of Huntington's disease.4 Spatiotemporal molecular medicine has a high scientific impact on investigation of cell–cell communication, microenvironmental function, and interactomes. The interaction networks of genes, proteins, and cells within the microenvironment play a decisive role in the production of chemoattractants for inflammatory cell recruitment and growth factor production for cancer cell extensions. For example, the liver microenvironment was considered as one of the easy locations for extrahepatic tumor metastasis5, 6 due to rich contents of acellular components for metastatic niche formation, special phenomic profiles of cellular components and cell-derived exosomes for the communication between resident cells, immigrated cancer cells and cancer stem cells, and dynamic processes of epithelial-to-mesenchymal or mesenchymal-to-epithelial transitions. With the rapid development of spatiotemporal measurements and analyses, we can define and monitor spatiotemporal gradient formation of acellular components, dynamic intra-/intercellular communications and regulations of cell signals, and dynamics of cancer stem cell heterogeneity and intratumoral immune cell landscape. In addition to spatial transcriptomics and proteomics, spatiotemporal microenvironment can be investigated by spatiotemporal target-paneled multi-omics, three-dimensional in vitro model of cell-engineered microenvironment, and spatiotemporal nanoparticle transitions. Spatiotemporal molecular medicine has the special power to explore molecular mechanisms and modifications of embryonic development, tissue morphogenesis, and organogenesis. Genomic dimensions, architectures, and organization play critical roles in regulations of gene transcription, spatiotemporal development, genome assembly, and target-based therapy.7 The developmental dynamics and regulations of multiple foci/cell clusters and regeneration from the same or different clones can be geographically and temporally dependent upon genomic characterization. The heterogeneity of genomic organizations exists in normal development and tumor growth/recurrence and is responsible for cancer cell responses to therapy. Lee et al. measured genomic and expression signatures of bulk cells, single cells, and selected regional cells isolated from tumor tissues and found the therapeutic response of glioma cells depended upon genetic similarity, mutations, and heterogeneities.8 Spatiotemporal molecular medicine could provide multidimensional insights into organogenesis during which gene regulations, protein interactions, and cell–cell communications in special positions are dynamically monitored for understanding mechanisms and prediction of diseases. It is possible to dynamically uncover the correlation of chromatin rearrangement with transcriptional networks, regulation of multiple factors at various levels to control cell fate and differentiation as well as rearrangement, and formation of organ-specific cell phenomes and function in specific cells. It allows for precise identification of the occurrence of dysfunction in signaling molecules, differentiation processes, cell types, and locations within the tissue architecture. From spatiotemporal morphogenesis, precision medicine will be further developed and improved for clinical diagnosis and therapy. Spatiotemporal medicine exists in clinic diagnosis, therapy, and prognosis for decades. The process of clinically examining responses of different organs to pathogens, diseases, and therapies at various stages and durations per se is a spatiotemporal course. Spatiotemporal images of organ can be constructed using ultrasounds, computed tomography (CT), nuclear magnetic resonance/CT, positron-emission tomography/CT, or pathological sections. Of those, four-dimensional ultrasound using spatiotemporal image correlation with software allows for detection of organ anatomy and function during aging and has special impact in disease diagnosis. Fluorescence endomicroscopy with automated image analysis can produce large quantities of data on the spatiotemporal behavior of target gene expression in cells of human digestive and respiratory tissues to analyze the geometry and fluorescence image and monitor temporal changes in gene expression. The temporal notion of causality between clinical phenomes observed over a period of time (e.g., symptoms, signs, images, biochemical measures, and responses to therapy) becomes an important part of spatiotemporal medicine, as suggested by considering shorter and longer time-series.9 Spatiotemporally fractionated treatment plans of photon radiotherapy for liver cancer were suggested as an efficient approach to save healthy tissue from the therapeutic target area.10 Spatiotemporal epidemiology has been also applied for monitoring the frequency of the diseases, for example, glanders, mycorrhizal fungi, Ebola virus spillover, as well as environmental and socioeconomic factors. Spatiotemporal molecular medicine contains an additional molecular level on top of spatiotemporal medicine and becomes appliable in clinical practice and investigation with the development of molecular barcode labeling technologies. Advanced methods to detect the spatial positioning of mRNA and protein expression in specific cells in tissue sections provide a new opportunity for defining the exact locations of disease-, disfunction-, signal-, and regulation-specific alterations in pathological conditions. This makes it possible to identify DNA methylation and transcriptomic information with protein expression in the immunophenotyping of cells on pathological images of ultrasounds and CT in the future, although many challenges remain to be overcome. The integration of spatiotemporal medicine with molecular spatial-omics maps can provide new strategies for early diagnosis and individualized therapies at molecular levels. The methodologies to merge image profiles and clinical phenomes with molecular profiles include in situ hybridization/fluorescence staining with oligonucleotide probes, digital barcodes tag with oligonucleotides, DNA microscopy with chemical DNA reactions fluorescent barcoding with sequencing, Drop-seq technology integrating single-cell RNA sequencing and imaging, padlock amplification of gene sequencing, cryogenic tissue sections with RNA sequencing and spatial data, cryogenic tissues from laser capture microdissection and RNA sequencing, cDNA synthesis in situ with spatial barcoding and RNA sequencing, mRNA barcoded in situ with spatial sequencing by oligonucleotide ligation and detection, virtual reconstruction of the tissue from single-cell RNA sequencing, and new transcriptome alkylation-dependent single-cell RNA sequencing to identify temporal and spatial features of single-cell data.11-13 The methods of spatial proteome represented by localizations and dynamics of proteins at subcellular level are potentials for global comparative applications to figure out cell variations, dynamic protein translocations, interaction networks, and compartments of protein location.14 Spatiotemporal molecular medicine is a part of molecular medicine characterized by gene-based diagnosis and therapy. With the improvement of on-/off-target site specificity and efficiency and improved understanding of structures, mechanisms, clinical applications, and off-target activities of genome editing systems, genome editing will become one of the clinical precision medicine strategies and multidisciplinary therapy strategies by integrating gene sequencing, clinical trans-omics, and single-cell biomedicine.15 As parts of spatiotemporal molecular medicine, spatial transcriptomics, proteomics, metabolomics, and bioinformatics can be the important tools to evaluate whether the body is suitable for targeted gene editing by programmable endonucleases prior to the treatment and monitor how the systems’ responses occur and how spatiotemporal controls might regulate functional switches or be regulated by gene editing associated alterations after the gene therapy.16, 17 Spatiotemporal molecular medicine requires the obvious contributions from rapid development of artificial intelligence, automatic robots, and computational and mathematical models. For example, the artificial intelligent cell was proposed as a system with computerized databases, digitalized informatics of biological elements, and programmed function and signals at a single-cell level.18 Although the system is still under development, spatiotemporal artificial intelligent cells with deep learning and auto-programming capacities, especially integrative capacity of clinical phenomic profiles, will make clinical application of spatiotemporal molecular medicine much easier. One of spatiotemporal molecular medicine aims is to translate spatiotemporal distributions of gene and protein profiles, regulations, and intercellular interaction and communication into clinical phenotypes, patient response to therapies, and strategies of precision medicine. The integration of single-molecule spectroscopy, multi-omic profiles, clinical phenomes, and numerical modeling is an approach to understand fluctuations of cell/organ responses and gene regulatory process from multiple single-cell datasets. Many computerized models have been developed, including air pollution spatiotemporal model for monitoring public health, spatiotemporal chromatin fluidity model for detecting heterogeneity dynamics, population shape decoding model for neuron spatio-temporal features, and network-based bioinformatics model for spatiotemporal RNA sequencing data analyses. Spatiotemporal computerization of human organ anatomy and function for medical education and clinical operation is still in the process of maturing. Spatial navigation and reinforcement learning of the brain are modeled using descriptive, mechanistic, and normative approaches assisted by artificial intelligence, which enables spatial positioning of pathological focus, correspondence between neuro-structure and behavior, and definition of molecular networks in neurons.19 As a new discipline, spatiotemporal molecular medicine faces many challenges, for example, how to clearly define and understand concepts, categorize contents, and standardize protocols of measurements and analyses. In conclusion, we propose the concept of spatiotemporal molecular medicine as a new discipline of medicine and believe it will become more important and applicable with advanced development of spatial and temporal transcriptomics, proteomics, and metabolomics. Clinical spatialization and temporalization of phenomes are integrated with the information from spatiotemporal molecular omics to form a four-dimensional understanding, diagnosis, and therapy for patients. Spatiotemporal measurements and analyses are decisive factors that may limit clinical applications of spatiotemporal molecular medicine. Although many challenges need to be overcome, spatiotemporal molecular medicine will improve our knowledge of diseases, provide multidimensional and targeted panels for more precise and early diagnosis and therapy, and be one of the future topics in clinical and translational medicine. The work was supported by Operation Funding of Shanghai Institute of Clinical Bioinformatics and Shanghai Engineering and Technology Center for Artificial Intelligence of Lung and Heart Diseases from Zhongshan Hospital, The National Nature Science Foundation of China (81873409), and National Key Research and Development Program of Precision Medicine (2017YFC0909500).

‘There is an urgent need to develop individuals who have the ability to combine a firm grounding in the principles of basic and clinical pharmacology with the most modern research technologies to address complex (patho)physiological questions’ Those who remember their school Latin will recall the paradigm for regular verbs such as ‘to love’: amo, amare, amavi, amatum – I love (present tense), to love (present infinitive), I have loved (perfect tense), loved (past participle). But the paradigm for the verb ‘to carry’ is irregular: ferro, ferre, tuli, latum. It comes, in fact, from three different roots. And that explains the connection between words such as refer and relate, confer and collate, transfer and translate [1]. To transfer means to carry over or bring over, and by extension to change over. To translate carries a similar meaning, to change or adapt to another use. This meaning, which dates from the late 14th century, includes such implications as physical or metaphorical transference and the interchangeability of languages, and it extends the meaning of transfer. ‘Bottom, thou art translated,’ says Snug in A Midsummer Night's Dream; not only has Bottom been changed physically into an ass; his metaphorically asinine nature has also been translated. These two ideas of carrying over and changing or adapting merge in the term ‘translational medicine’, which was first formulated in the mid 1990s and caught on quickly enough for a journal, The Journal of Translational Medicine, to be founded in 2003 [2]. Other non-identical but overlapping terms, such as ‘systems biology’ and ‘experimental medicine’ or ‘experimental therapeutics’, have also emerged in an attempt to cover the spectrum of what is increasingly being alliteratively referred to as science ‘from bench to bedside’ and ‘from molecules to man’. However, like so many apparently new ideas, this one is as old as the hills – only the terms are new. Clinical academics have been interested in the two-way street that links basic science and clinical applications ever since clinical research was invented. And clinical pharmacology is the subject that best exemplifies this idea: its activities encompass everything from the molecular end of drug therapy to the use of medicines in individuals and populations (Figure 1) [3]. Clinical pharmacological tools for translational medicine (adapted from [3]) In this short review we cannot cover all the aspects of the tools of clinical pharmacology that make it relevant to translational medicine; we shall simply highlight a few examples. The publication of the sequence and initial analysis of the human genome in 2001 [4] fuelled media hype and high expectations concerning ‘personalized medicine’. In the early 21st century this idea has become closely linked to pharmacogenetics and pharmaco-genomics [5], although clinical pharmacologists, using non-genetic as well as genetic tools, have always espoused the concept of individualized drug therapy, even though it is a holy grail that cannot always be obtained (for example, in mass strategies such as immunization). Moreover, individual variability in drug response depends on non-genetic and environmental factors, in addition to genotype [6]. But pharmacogenomics is potentially a major method in the development of translational medicine and is a current buzz-word in biomedical research and research funding, as demonstrated by the MRC-NIHR Efficacy and Mechanism Evaluation (EME) Programme in the UK, due to be launched later this year [7], and the NIH Clinical and Translational Science Awards in the USA [8]. Pharmacogenomic tools for translational medicine were well represented in the Journal during 2007. Dickinson et al. [9] used a mechanistic population-based pharmacokinetic (PK) and pharmacodynamic (PD) model to define the size and power of in vivo studies needed to discern the effect of genotype on PK and PD, using variants of the CYP2C9 genotype (*2/*2, *2/*3, and *3/*3) and (S)-warfarin as a model. Their modelling suggested that about 250 subjects were needed to detect a difference in anticoagulant response between genotypes. Experimental PK and PD comparisons between wild-type and other individual CYP2C9 genotypes showed that only 21% of cases (20 of 95 comparisons within 11 PD and four PK-PD studies) reported statistically significant differences. Simulations of studies enriched with specific genotypes showed that only three and five subjects were required to detect differences in PK and PD between the wild type and the *3/*3 genotype. An extension of these sophisticated studies would add information about genetic variants of vitamin K epoxide reductase complex 1 (VKORC1 –the pharmacological target of warfarin action), and such analyses could be used to design and power more efficient clinical studies of the problem of variability in anticoagulant responses. The UGT1A1*28 polymorphism of UDP-glucuronyltransferase reduces UGT1A1 enzyme activity, via an extra TA repeat in the promoter region. During 2006 the FDA modified the drug labelling of the cytotoxic drug irinotecan (CPT-11) and suggested that genotyping for this polymorphism could assist in guiding irinotecan dosing, thus avoiding serious neutropenia [10]. Peterkin et al. measured UGT1A1 enzyme activity with probe substrates and protein expression levels ex vivo in human hepatic microsomes [11]. Their results suggested that the UGT1A1*28 polymorphism contributes about 40% of the variability in UGT1A1 enzyme activity. This is in keeping with clinical data that suggest that UGT1A1*28 is more important when using high-dose irinotecan [12]. This study and that of Dickinson et al. suggest that to define better the extent to which genotype contributes to individual variability in drug responses we must study the genetic variability in the proteins that are involved in the drug pathways (targets, drug transporters, metabolizing enzymes). Such genetic information is being compiled, and several drug pathways have been described in the PharmGKB pharmacogenetics and pharmacogenomics knowledge database [13]. The use of such information should lead to meticulously planned clinical studies to test how this genetic knowledge can guide optimization of drug therapy. In our sister journal, The British Journal of Pharmacology, Rognan introduced and comprehensively reviewed chemogenomics [14]. This discipline systematically investigates the biological effect of a spectrum of small ligands on a wide array of macromolecular targets. Because the quantity of existing data (compounds, targets, and assays) and of the resulting information (gene/protein expression levels and binding constants) is too large for manual manipulation, information technology plays a crucial role in planning, analysing, and predicting chemogenomic data. These methods permit navigation of either ligand or target space and should lead to more scientifically driven drug discovery paradigms. In order to translate further advances in the fields of pharmacogenomics (as ‘personalized medicine’) or chemogenomics (in the clinical development of novel therapeutics), a multi-disciplinary team, with clinical pharmacologists in the forefront as experts in translational techniques, is a prerequisite. Biomarkers are essential tools in translational medicine [15] (Figure 1), and examples appear in the Journal from time to time [16]. During 2007, studies of different types of new potential biomarkers included plasma deoxynucleoside concentrations in cancer chemotherapy with capecitabine [17], the flurbiprofen urinary metabolite ratio as an in vivo measure of CYP2C9 activity [18], and metabolite profiles as biomarkers for the pharmacological effects of thiazolidinediones in diabetes mellitus [19]. Other studies involved the use of biomarkers that are already well known, such as arterial stiffness in hypertension [20], prolactin secretion as a measure of dopaminergic function [21], and cortisol and ACTH as markers of the effects of exogenous glucocorticoids [22]. Not all of the new potential biomarkers will prove useful in predicting outcomes, but even if they do not, such studies add to our cumulative understanding of the pharmacology of the relevant drugs. In 2007 the Journal published several papers dealing with other types of new techniques. For example, Ndovi et al. described an intriguing method for measuring drug concentrations in the male genital tract using a device for collecting split fractions of the ejaculate in healthy men [23]; van der Schueren et al. described a technique for studying capsaicin-induced neurogenic inflammation [24]; and Drummond studied the effect of electrical stimulation on the axon reflex [25]. These methods are all of potential use in translational medicine. Indeed, the Editors of the Journal have recognized the importance of such studies by awarding the 2007 BJCP Prize to Dr Ndovi*, jointly with two other trainees in clinical pharmacology. These papers demonstrate the power of clinical pharmacological methods as translational tools. The British Journal of Clinical Pharmacology is not The British Journal of Translational Medicine, because clinical pharmacology is more than translational medicine. But much of translational medicine will be served by clinical pharmacologists and the tools that they provide. This is reflected in many of the articles that appear in the Journal, only a few of which we have highlighted here. The recent initiative of the Wellcome Trust in establishing interdisciplinary training programmes for clinicians in translational medicine and therapeutics [26] emphasizes that as translational medicine develops it will continue to depend heavily on the methods of clinical pharmacology and the expertise of its practitioners. The opening paragraph of the Trust's call for applications says it all: ‘There is an urgent need to develop individuals who have the ability to combine a firm grounding in the principles of basic and clinical pharmacology with the most modern research technologies to address complex (patho)physiological questions. Such individuals will play a key role in shaping the interdisciplinary research that underpins translational medicine and therapeutics.’ We concur and could not have put it better.

The global scientific community is going to blast off to the Moon again with the clear aim to cure cancer 47 years after man first stepped onto the lunar surface. President Obama signed a Presidential Memorandum at the beginning of 2016—White House Cancer Moonshot Task Force—and called upon the world to “cure cancer once and for all”. It is a new milestone and effort to re-unite scientific resources and experts, optimize therapeutic strategies and improve outcomes of patients with cancer. The Cancer Moonshot initiative delivers a number of opportunities and challenges, e.g., how can we integrate the Moonshot program with other anti-cancer strategies like clinical and translational medicine or precision medicine; how do we define focus points; how do we monitor progress along the “Moonshot” voyage? The cancer research community is ready for the Moonshot 2020 program, although there are still a large number of thresholds, hurdles and obstacles to be determined and resolved along the way. A number of advanced biotechnologies in genomics, proteomics, metabolomics, systems biology, clinical bioinformatics and drug discovery have been translated into clinical practices and applications. Great potential lies in single-cell biology as well as in the study of the 3D architecture and organization of the genome, also known as gene repositioning. New therapies for gene editing, cell targeting, and immune approaches give the Cancer Moonshot 2020 program the ammunition required to reach its goals. For example, preclinical studies have shown the potential for multiplexed genome editing in in vitro and in vivo systems by CRISPR/Cas-mediated gene editing. Wang et al. simultaneously interrupted five distinct alleles in mouse embryonic stem cells, with an efficiency of 80 % to achieve the one-step generation of animals carrying mutations in multiple genes [1]. Transcription activator-like effector nucleases (TALENs) and clustered regularly interspaced short palindromic repeats (CRISPR)–Cas9 nucleases have been suggested as the most commonly employed approaches to edit the human genome. Another exciting development is the application of antibodies in the inhibition of endogenous immune responses to cancer, known as checkpoint blockade therapy. Clinical trials have proved the safety and efficacy of antibodies that block the T cell inhibitory molecules CTLA-4 and PD-1 in treating subsets of patients with metastatic melanoma and renal cell carcinoma [2]. There is however quite some discussion about the types of cancer that would be most suitable for intervention using immunotherapy and there is great interest in specific biomarkers to monitor the various responses to these therapies. We should also consider the potential for overlap and cooperation between the Cancer Moonshot 2020 program and clinical and translational medicine, and precision medicine. One part of the Cancer Moonshot 2020 program is to use progress in clinical and translational medicine, to re-unite, re-collect, re-organize, and re-optimize resources to fight diseases [3, 4]. Clinical and translational medicine was initiated 10 years ago to foster the communication between basic and clinical scientists, to translate advanced biotechnologies into clinical practice, and to improve the quality of life for patients. Precision medicine was announced in 2015 as a new emerging area and therapeutic strategy to improve the treatment and prognosis of patients by integrating clinical phenotypes with bioinformatics, computational science, mathematics, gene sequencing and systems biology. Precision medicine was proposed, containing five critical elements: clinical bioinformatics, precision methodologies, disease-specific biomarkers, drug discovery and development, and precision regulations to guard the application of precision medicine [5]. A gene-based therapeutic strategy was initially suggested for inherited diseases and cancers on the basis of gene mutation and heterogeneity, and it was then later extended to include metabolic and cardiovascular diseases. Precision medicine could be one of the therapeutic strategies in the program of Cancer Moonshot 2020 program, or the program could be one of the therapeutic areas of focus in precision medicine. The “White House Cancer Moonshot Task Force” was defined as a national program, government-led commission, Federal investments-dominated research, and cancer-focused therapy [7]. However, the fight against cancer is a global issue, which will greatly benefit from the involvement and contribution of the entire scientific community, focusing on a clear global strategy. To be able to effectively tackle the issues ahead, we need to develop a better understanding of cellular structures, organelles, functioning properties, metabolic circles, signaling pathways or interactions, and of the tumor microenvironment. We should identify molecular targets during the discovery and development of target-based drugs and have biological function- and disease-specific biomarkers to monitor drug efficacy, efficiency and toxicology. We should have a better understanding of alterations of spatial genome organization and its effect on transcription as well as of higher-order chromosome folding and specific chromatin interactions in the topological mechanism of human cancer [8]. The Cancer Moonshot is a new ‘giant leap’ for clinical and translational medicine, a new opportunity to optimize scientific achievements and advances to cure cancer, with a special focus on cancer prevention and early detection and treatment. The success of the Cancer Moonshot 2020 program will also be dependent upon the sharing of databases, which include clinical information, patient phenotypes, imaging, biochemical measurements, therapies, gene sequencing and systems biology. We therefore believe that clinical and translational medicine will play an important role in the achievement of the goals set out in the Cancer Moonshot 2020 program. Cancer Moonshot “Apollo 11” has been launched and the anti-cancer spacecraft have started their voyage; the countdown to cure cancer has begun.

‘The Ultimate Computer’, an episode of the legendary 1968 series Star Trek: The Original Series, in retrospect, put the spotlight on the unprecedented development of artificial intelligence (AI), with its enormous possibilities, but also limitations and dangers, without knowledge of the modern computational methods of the 21st century. I recently watched this episode under this premise and was stunned by the conversations about the benefits and dangers of artificial intelligence that were already taking place at that point. In 2011, this episode was named by Forbes as one of the 10 best episodes of the series, dealing with the effects of advanced technology and, in particular, the dangers of artificial intelligence.1 Forbes commented: ‘The Ultimate Computer’ (TOS) - This is a pretty fascinating episode, especially considering the time period. Pre-dating HAL by a few years, this episode features M-5, a computer capable of learning and adapting that is given command of the Enterprise on a test run. Naturally, things don't go well, but it's an interesting exploration of the difference between human and machine intelligence. It also serves as a metaphor for people who lose their jobs due to mechanization’. “ChatGPT, a state-of-the-art language model developed by OpenAI, has the potential to revolutionize the way medical information is disseminated and processed in the clinical and translational medicine fields. Here are some of the key benefits, challenges, pitfalls and future perspectives of using ChatGPT in this context”: Access to up-to-date information: ChatGPT can access vast amounts of medical data and provide real-time, accurate answers to clinical questions. Improved patient engagement: ChatGPT can provide patients with a convenient and accessible way to get answers to their medical questions and to track their health information.1 Reduced workload for healthcare providers: ChatGPT can help reduce the administrative burden on healthcare providers, allowing them to focus more on patient care. Data privacy concerns: ChatGPT may raise privacy concerns as it has access to vast amounts of personal health information. Inconsistent accuracy: ChatGPT may not always provide accurate answers, particularly for complex medical questions. Bias in the training data: ChatGPT is only as good as the data it was trained on. If the training data are biased, the model may perpetuate that bias. Misleading information: ChatGPT is not a substitute for a healthcare provider, and incorrect or misleading information may cause harm to patients. Dependence on technology: ChatGPT may become a crutch for healthcare providers, reducing their ability to diagnose and treat patients without relying on technology. Improved accuracy: As more data are collected and analyzed, ChatGPT's accuracy is likely to improve. Integration with electronic health records (EHRs): ChatGPT has the potential to be integrated with EHRs, allowing for a more seamless exchange of information between healthcare providers and patients. Personalized medicine: ChatGPT has the potential to provide personalized medical advice based on individual patient data and health histories. In conclusion, ChatGPT has the potential to significantly impact the clinical and translational medicine fields by providing access to up-to-date information, improving patient engagement and reducing workloads for healthcare providers. However, there are also challenges and pitfalls to be considered and addressed, and continued research and development are needed to ensure that ChatGPT is used safely and effectively’. When repeating the same query (“Regenerate Response”), it is interesting to note that the structure of the text remains unchanged, but with some deviations and variations from the previous result (see Figure 1). These differences can be explained by the underlying structure and learning architecture as well as the parameter settings of the tool, making it difficult to reproduce repeated queries. One should know that ChatGPT is deep learning model that uses a transformer-based neural network to generate human-like text. It is intriguing that the original question about the potential impact of ChatGPT in clinical and translational medicine seems to be almost answered by ChatGPT itself, yet there are some questions that need to be addressed here as a result of the global discussion that has begun about ChatGPT and its impact in healthcare. There is no doubt that the benefits and advantages of AI in medicine are beyond question, and have already made their way into everything from research to clinical applications. AI-based algorithms in all areas from data-driven preclinical research to medical decision support in the daily clinical practice, for example, through AI-based interpretation of medical images for diagnosis and patient management, pattern recognition and pre-interpretation of physiological and biophysical signals such as ECG, EEG, EMG and others, are already parts of standardized medical procedures according to GCP. However, the challenges and pitfalls indicated by ChatGPT naturally raise many questions, including relevant medical ethics issues. Data privacy and security, incorrect or misleading information that may cause harm to patients. Otherwise, patients are well prepared for the doctor's interview and appear confident on relevant clinical information such as disease-specific symptoms, medical history, and so on. However, this is often not in the interest of the attending physician and can lead to misunderstandings. In addition, the inconsistent accuracy of ChatGPT responses can lead to problems in education if ChatGPT is used without authorization. Some universities have already banned this tool. Medical staff in training and students could be influenced accordingly by ChatGPT and misinterpret medical knowledge or even make clinical decisions without independent and critical evaluation and validation. The most critical technical issue is training data bias, as ChatGPT's underlying algorithms are only as good as the data they are trained on, and are therefore prone to errors in their responses. Consequently, it would be part of GCP to train and validate the ChatGPT algorithm, for example, for diagnosis and therapy-accompanying applications on relevant, evidence-based knowledge bases, before it can be used. Here, approaches and concepts as known from medical device approval should be considered, including requirements and usability engineering, risk assessment and clinical evaluation. In summary, there are many other aspects that could be discussed in this commentary to illustrate the tremendous benefits of open AI, while not ignoring the consequences and implications of using ChatGPT unwisely. Returning to the introductory example of the science fiction episode ‘The Ultimate Computer’, the advantages and pitfalls of an AI-controlled world were impressively portrayed in this story and have also proven to be real to this day. With the launch of the OpenAI project in 2022, the world has finally been catapulted into a new ‘age of artificial intelligence’, and humans are being called upon to take ultimate responsibility for these new technologies. ‘ChatGPT has the potential to significantly impact the clinical and translational medicine fields by providing access to up-to-date information, improving patient engagement and reducing workloads for healthcare providers. However, there are also challenges and pitfalls to be considered and addressed, and continued research and development are needed to ensure that ChatGPT is used safely and effectively’. In the hope and ethical obligation to improve the quality of life of patients. The authors declare no conflicts of interest.

The Chat Generative Pre-trained Transformer (ChatGPT) is an artificial intelligence (AI) model developed by Open AI for generating human-like text. ChatGPT is powered by the large language model (LLM) GPT-3.5. Like other LLM-based AIs, it has been trained on large datasets of text and can generate new text similar to the text it was trained on, which requires generating, understanding and interpreting human language using computer systems.1-3 ChatGPT simulates human interaction, and it is probably the most powerful language model currently available. It uses a deep learning technique called a ‘transformer architecture’ that is trained on massive text datasets obtained from the internet.4 ChatGPT is leading a revolution in AI technology, and its impact on the field of clinical medicine is yet to be determined. Some clinical practices may rely on data analysis, clinical research and guidelines. AI models may help in clinical decision support, clinical trial recruitment, clinical data management, research support, patient education and other fields.5, 6 In some cases, AI models can automate certain tasks performed by humans, such as data analysis, image acquisition and interpretation.7, 8 This may increase the efficiency and reduce the workload of healthcare professionals, allowing them to focus on higher level tasks that require their expertise and clinical judgement. On the other hand, the results must originate from the extraordinary human mind and its ability to process information and communicate through them.9 Therefore, it is important to use AI models such as ChatGPT as a tool to support, rather than replace, healthcare professionals in their decision-making process. Similarly, AI technology including ChatGPT also has great potential in assisting basic research and accelerating the technological transformation of clinical and translational medicine. For example, in terms of drug discovery, AI has image recognition capabilities, which can identify, classify and describe chemical formulas or molecular structures to assist the design of new structures and functional group combinations of compounds.10 In addition, AI technologies play an important role in disease prediction, diagnosis and assessment of therapeutic targets, such as providing treatment guidelines for cancer patients based on their magnetic resonance imaging radiomics and predicting ageing-related diseases.11-13 However, unlike the AI algorithm or model specially developed for drug discovery and disease diagnosis, the core value and advantage of ChatGPT lies in its powerful LLM. At present, ChatGPT cannot update the training data in a real-time manner. In addition, it can only give general and vague answers in some existing medical-related conversations.14 Some experts present conversations with ChatGPT online for case studies and they find that the diagnoses made by ChatGPT are often not comprehensive and adequate. For example, for patients with common symptoms such as fever, ChatGPT will give suggestions to take antipyretics to help relieve symptoms, but cannot accurately judge infection, pimples or other causes. Therefore, blindly relying on the diagnosis and guidance of ChatGPT may have the potential risk of inaccurate diagnosis or delayed treatment. Furthermore, Fijačko et al. pointed out that ChatGPT without specific courses or training in medical knowledge could not pass the life-support exam, suggesting that ChatGPT may not be capable enough for life support in clinical applications.15 This evidence implies that ChatGPT may not be capable of independently handling the complex work of clinical practice in the current version. Therefore, focusing on the application of ChatGPT in human–computer interaction may improve its usability in clinical practice. For example, the diagnosis and treatment of mental illness rely heavily on doctor–patient questionnaires, interviews and judgement. However, many interfering factors such as the physician's tone of voice, mood and the surrounding environment can hinder an accurate assessment of the disease. In this field, AI has been used to record and analyze data related to questionnaires.16 The emergence of ChatGPT may accelerate the integration of flexible questionnaires, documentation, diagnosis and follow-up of patients with mental disorders using chatbots. In addition, ChatGPT provides a basis for more flexible and efficient epidemiological research. Indeed, epidemiological research also relies on efficient and reliable data collection, recording and analysis.17 ChatGPT not only solves the difficulty of remote inquiries, but also helps to reduce labour requirements to complete the work. Besides, ChatGPT is usually more accurate and faster than manual statistics and records. ChatGPT has caused some changes and controversies in a short period of time in terms of medical education, training and writing.18, 19 It has been used to finish writing an abstract, introduction or even the main text of an assignment or article, but it may not do well in writing the creative part of medical writing.20 ChatGPT often summarizes previous research and data to form abstracts or background knowledge. However, some issues, including ethical concerns, need further clarification.21 Some scientific journals noted that articles containing ChatGPT-generated content need to be explicitly listed as authors, while Nature refused to accept ChatGPT as an author on articles because it cannot be responsible for content generated by itself.22 Furthermore, AI models including ChatGPT can clearly help the healthcare industry by providing a more objective and evidence-based approach to decision-making, reducing the risk of human error based on its unparalleled speed of information processing. It also helps identify patterns and correlations in vast amounts of data, which may provide new insights and discoveries in medicine. Additionally, AI models can assist in the detection of disease and the prediction of prognosis,23 which may help provide more personalized treatment recommendations and improve patient outcomes.24 Further research and development are needed to ensure the effective and responsible use of AI models in the healthcare industry. Nevertheless, we should be aware that ChatGPT is a double-edged sword, with both powerful features and potential shortcomings.3 AI has the potential to significantly impact clinical and translational medicine by improving data analysis,7 streamlining workflow25 and enhancing decision-making.26 However, potential negative impacts such as privacy concerns,27 bias, discrimination28 and so forth should not be underestimated. Overall, ChatGPT has the potential to revolutionize clinical and translational medicine, but we need to develop appropriate strategies to mitigate potential risks and negative outcomes. Despite our initial lack of preparation for the game-changing ChatGPT technology, the development of AI is unstoppable. The best course of action is to embrace it, use its capabilities to improve our lives, and foster mutually beneficial relationships by evolving it in clinical medicine.

Measurements, monitoring, and tuning of body processes such as muscle contraction, thermoregulation, circulation and digestion hold a significant place in the field of clinical and translational medicine. From pre-treatment evaluation and treatment effectiveness tracking to managing chronic diseases, crucial diagnostic information is provided from those measurements and monitoring via medical devices and systems.1, 2 Creative translational medical devices and systems play a vital role in advanced health care. From the materials perspective, semiconductors are an essential type for those advanced devices that come into play. For example, semiconductors are used in sensors, and imaging systems to monitor various physiological parameters. From the form factor perspective, fibre-shaped devices have long been used in medicine. Equipped with micro-optical fibres, the endoscope is advanced in providing high-resolution imaging and enhanced diagnostic capabilities. Catheters, usually made of polymeric fibres, are used in medicine for various purposes, including drainage, and administration of fluids and medications. Thus bringing the semiconductor materials in the fiber form factor closer together may reach the “best of both worlds” scenario. Conventional semiconductor materials such as silicon (Si) and germanium (Ge) are brittle inorganic crystalline materials, challenging their fabrication in fibre form. Our latest study unveiled that the thermally-drawn fibre technique can fabricate semiconductor fibre in high quality, high yield and extended single-strand length through thermomechanical optimization.3 From submicrons to hundreds of microns, the controllable thickness of these semiconductor fibres makes them suitable for use and integration into different kinds of medical devices. There's still much to learn on establishing and optimizing their connectivity to medical devices, semiconductor fibres, not expected to replace their planar-type counterparts, will certainly unlock new and exciting opportunities in clinical and translational medicine (Figure 1). Thermoelectric semiconductors convert heat into electricity (Seebeck effect) or vice versa (Peltier effect). This feature brings them applications in power generation, temperature sensing, and regulation. Our recent work delivered tin selenide fibres.4 In the fibre form factor, the thermoelectric semiconductor possesses remarkable characteristics such as being lightweight, thin, and mechanically flexible.4-6 Breathable functional fabric that offers long-term comfort can be prepared using these tin selenide fibres. The functional fabric measures body temperature and outputs electrical signals. A woven fibre crossbar structure offers multiple sensing nodes. Each node is created on the crossing of two fibres. 2N number of fibre can form N2 nodes, and thus, the functional fabric allows the monitoring of a large area covering the body. This functional fabric provides thermoregulation through the other working modes of thermoelectric semiconductors. Power input generates two sides, the hot and cold sides and polarity switching ensures that the side with skin contact is adjusted on demand. Photoplethysmography (PPG) is a technique that can measure heart rate, mainly using a light source and a photodetector. A small portion of the incident light is scattered back to the photodetector and generates an electrical signal. As the volume of blood vessels changes with the heartbeats, the electrical signal from the photodetector is modulated by the heart rate. A flat and nonconformal photodetector is usually installed for conventional fingertip PPG devices or smartwatches. The measurements are conducted with the testee remaining still to ensure decent contact between the photodetector and skin, which challenges continuous monitoring of such a physiological parameter. In our study, we employed semiconductor fibres to prepare a fibre photodetector, which is soft and flexible to be woven into a wrist strap. This functional strap is conformal to the wrist and is able to measure pulse through PPG. Due to its flexible nature, this sensor provides more accurate monitoring when the testee is active and is suitable for 24-h continuous monitoring. The study suggests that semiconductor fibres have diverse clinical and translational medicine capabilities. The potential of semiconductor materials in such a fibre form factor lies in various applications such as implantable sensors, neural interfaces and tissue engineering. The form factor of semiconductor fibres promotes their integration in catheters and other devices for insertion. As one of many examples, a waveguide of extended wavelength can be integrated into an endoscope using semiconductor fibres. Also, they can be used as neural probes for interfacing with the nervous system. As part of the invasive neural probe, semiconductor fibres can be used to record neural signals, stimulate neural activity either by electrical or optical signals and establish bidirectional communication between the brain or spinal cord and external terminals. Furthermore, biocompatible and bioresorbable semiconductor fibres can be utilized as scaffolds for tissue engineering and regenerative medicine. They might be employed as mechanical support and through stimuli or medicine delivery to promote tissue repair and regeneration. Nevertheless, more understanding of the biophysics of semiconductor fibres will be needed for practical evaluation of their use in clinical usefulness. We hope scientific breakthroughs and advanced technologies continue to emerge and improve in the field to unlock the clinical roles and potential therapeutic benefits for semiconductor fibres. All authors have contributed to writing the manuscript and have approved the final manuscript. This work was supported by the Singapore Ministry of Education Academic Research Fund Tier 2 (MOE2019-T2-2-127, MOE-T2EP50120-0002 and MOE-T2EP50123-0014), the Singapore Ministry of Education Academic Research Fund Tier 1 (RG62/22), A*STAR under AME IRG (A2083c0062), A*STAR under IAF-ICP Programme I2001E0067 and the Schaeffler Hub for Advanced Research at NTU, the IDMxS (Institute for Digital Molecular Analytics and Science) by the Singapore Ministry of Education under the Research Centres of Excellence scheme, and the NTU-PSL Joint Lab collaboration. The authors declare no conflicts of interest. Not applicable.

ChatGPT, an artificial intelligence (AI)-powered chatbot developed by OpenAI, is creating a buzz across all occupational sectors. Its name comes from its basis in the Generative Pretrained Transformer (GPT) language model. ChatGPT's most promising feature is its ability to offer human-like responses to text input using deep learning techniques at a level far superior to any other AI model. Its rapid integration in various industries signals the public's burgeoning reliance on AI technology. Thus, it is essential to critically evaluate ChatGPT's potential impacts on academic clinical and translational medicine research. ChatGPT contains 175 billion parameters, making it one of the largest and most powerful models for AI processing available today—hence its growing use in different occupations. ChatGPT's responses are leaps and bounds above those from past AI programs, in no small part due to being more human-like. ChatGPT has taken the business world by storm. It is easy to envision its expansion into clinical and translational medicine in the future. As such, experts must consider the potential effects of this technology in and beyond medical research. ChatGPT has made its debut in the scientific literature through published papers and preprints. Although ChatGPT can undoubtedly benefit writers of all backgrounds, its limitations in medical research merit close attention.1 The emerging use of ChatGPT has sparked an upheaval in the scientific community and ignited debates around the ethics of using AI to write scientific publications that can influence the decisions of physicians, researchers, and policymakers. The most significant disadvantage of ChatGPT is that the information it compiles is not always accurate. This drawback is especially detrimental in academic publishing; after all, progress depends on sharing appropriate information. Presenting incorrect data in a scientific setting carries a great risk of harm. For example, research influences how personal and community health concerns are treated and managed. The data which ChatGPT uses provide information from 2021 and earlier. The chatbot does not currently consider information reported in 2022 onward.2 For a field that is driven by recent advances to boost knowledge, enhance interventions, and formulate evidence-based policies, this year-long (and growing) information gap is a stark hindrance. If scholars use ChatGPT to create content, attempting to publish papers that contain false or outdated information will tarnish authors’ credibility among colleagues and peers. A double-edged sword with ChatGPT is the ability—or more accurately, the inability—of scholars to detect when other professionals have used it. Researchers at Northwestern University asked ChatGPT to write 50 medical-research abstracts based on a set of articles published in medical journals. The authors then asked a group of medical researchers to spot the fabricated abstracts.3 Problematic results emerged, with human reviewers able to correctly identify only 68% of the ChatGPT-produced abstracts and 86% of the genuine abstracts. These findings confirm ChatGPT writes believable (albeit potentially inaccurate) scientific abstracts. The results of this study bode well for those interested in employing ChatGPT to facilitate the writing process, as people reading their work likely will not realize it was AI-generated. However, this possibility raises several concerns. Being unable to identify valid information comes with consequences. Scientists may follow flawed investigation routes, which translate into wasted research dollars and misleading results. For policymakers, the inability to detect false research may ground policy decisions in incorrect information that could have monumental effects on society. Due to these implications, the future of academic and scientific publishing may soon hold policies that forbid AI-generated content. Those who use ChatGPT in any capacity will need to be aware of these mandates. The 40th International Conference on Machine Learning already banned papers written by AI tools, including ChatGPT.4 The Science family of journals is also updating their license and Editorial Policies to specify that they will not allow ChatGPT-produced text. They explained their stance in an editorial, stating that most cases of scientific misconduct arise from inadequate human attention, and permitting ChatGPT-generated content significantly increases this risk.5 Not all ChatGPT-related matters have elicited concern within the scientific research field. A February 2023 article in Nature described computational biologists’ use of ChatGPT to improve completed research papers. In just five minutes, the biologists received a review of their manuscript that increased readability and spotted equation-based mistakes. During a trial with three manuscripts, the team's use of ChatGPT was not always smooth, but the final output returned better-edited manuscripts.6 Using ChatGPT for this purpose bypasses the scientific community's primary concerns surrounding AI and its use of inaccurate or outdated information. Because computational biologists initially wrote the manuscripts, the information was already accurate and up to date. ChatGPT can help increase researchers’ productivity and content quality. If scientists can spend less time editing their work, they can devote more time to advancing the field of medicine. Considering these benefits, ChatGPT can prove invaluable for researchers looking to verify answers or identify problems in their work. It is important to remember that, as of now, ChatGPT is not sufficiently trained on specialized content to be able to fact-check technical topics.7 Experts anticipate that the technology and programs integrating ChatGPT will serve as precursors to more advanced AI systems. In the meantime, this chatbot can play a supportive role in academic and scientific publishing, primarily for editing. Even so, those who use ChatGPT must be aware of its limitations. As it stands, ChatGPT cannot be relied upon to provide correct facts or produce reliable references, as stated by a January editorial in Nature Machine Intelligence.8 Accepting the limitations of ChatGPT and using it only for certain tasks allows researchers to delegate tedious jobs, such as manuscript editing, to the AI model while avoiding catastrophes such as the publication of false information. As ChatGPT becomes more commonplace, it will be crucial to calibrate expectations about its capabilities and acknowledge that it cannot take on every job. Especially in the academic research field, any tasks in need of specialized subject knowledge or innovative ideas and opinions still require a genuine human touch that cannot be replaced by AI. Our conclusions regarding ChatGPT and its applications in scientific research focus on a high-impact journal–Clinical and Translational Medicine–that aims to promote, accelerate, and translate preclinical research for clinical applications. This journal highlights the importance of clinical and translational medicine research in the name of promoting the safety and efficacy of discoveries that proceed to human trials, reflecting the notion of ‘bench to bedside.’9 Implementing ChatGPT in its present iteration must be pursued with extreme caution given the tool's evolving limitations and capabilities when it comes to providing reliable information. Can AI replace human input? We concur with H. Holden Thorp's position5 on ChatGPT in that “ChatGPT is fun, but not an author” (p. 313). Scientists might be able to use well-developed AI tools to increase work efficiency for tasks such as proofreading and manuscript checks. In the future, AI-based tools may become recognized for their contributions to broader areas of scientific research, depending on their abilities to support human input. The boundaries between research ethics and the moral use of AI in health research10 need to be further explored to establish guidelines. All researchers and contributors must understand what AI can and cannot do. Therefore, editors and editorial board members should continue monitoring ChatGPT's applications in academic research to draft journal policies that inform contributors of best practices. Doing so will ensure that Clinical and Translational Medicine can maintain an image of integrity by publishing timely and accurate research that makes meaningful contributions. After all, research excellence is gauged by ethics and integrity. The authors declare no conflict of interest.

With the rapid development of biotechnologies and deep improvement of knowledge, "Discovery" is the initial period and source of innovation of clinical and translational medicine. The international journal of Clinical and Translational Discovery serves to highlight unknown or unclear aspects of clinical and translational medicine-associated knowledge, technologies, mechanisms, and therapies (https://onlinelibrary.wiley.com/journal/27680622). The Discovery aims to define the interaction between genes, proteins, and cells, and explore molecular mechanisms of intercommunication and inter-regulation. More discoveries of technologies and equipment are expected to improve method sensitivity, specificity, stability, analysis, and clinical significance. The first priority of Clinical and Translational Discovery is to turn gene-, protein-, drug-, cell-, and interaction-based discoveries into health advancements. Clinical and Translational Discovery highly focuses on the discoveries of biological therapies and precision medicine-based therapy elicited from computational chemistry, DNA libraries, target-dependent small molecular drugs, high-throughput screening, vaccination, immune therapy, cell implantations, gene editing, and RNA- or protein-based inhibitors. Thus, Clinical and Translational Discovery sincerely welcome you to join and share the rapid development and future successes to come.

With rapid developments of single-cell sequencing and multi/trans-omics, clinical single-cell biomedicine is a new and emergent discipline to integrate single-cell molecular and clinical phenomes and uncover new disease-specific diagnoses and therapy. The journal of Clinical and Translational Medicine (CTM) launches the first CTM initiative of clinical single-cell biomedicine (cscBioMed) to promote the discovery and development of single-cell-based biology and medicine, speed the translation from single-cell biology into clinical application, and improve early diagnosis and therapy for human diseases. The cscBioMed initiative is speeding translational processes from circulating single-cell RNA sequencing into routine measures in clinical biochemistry of haematology, from spatial transcriptomics into single-cell pathology, and from single-cell-based biomarkers and targets into clinical diagnostics and target drugs. With a clear goal, we expect that cscBioMed will benefit human health by establishing a clinical single-cell dynamic monitoring and early predicting system and by improving diagnosis and treatment.

Objective: To describe the infrastructure, tools, and services developed at Stanford Medicine to maintain its data science ecosystem and research patient data repository for clinical and translational research. Materials and Methods: The data science ecosystem, dubbed the Stanford Data Science Resources (SDSR), includes infrastructure and tools to create, search, retrieve, and analyze patient data, as well as services for data deidentification, linkage, and processing to extract high-value information from healthcare IT systems. Data are made available via self-service and concierge access, on HIPAA compliant secure computing infrastructure supported by in-depth user training. Results: The Stanford Medicine Research Data Repository (STARR) functions as the SDSR data integration point, and includes electronic medical records, clinical images, text, bedside monitoring data and HL7 messages. SDSR tools include tools for electronic phenotyping, cohort building, and a search engine for patient timelines. The SDSR supports patient data collection, reproducible research, and teaching using healthcare data, and facilitates industry collaborations and large-scale observational studies. Discussion: Research patient data repositories and their underlying data science infrastructure are essential to realizing a learning health system and advancing the mission of academic medical centers. Challenges to maintaining the SDSR include ensuring sufficient financial support while providing researchers and clinicians with maximal access to data and digital infrastructure, balancing tool development with user training, and supporting the diverse needs of users. Conclusion: Our experience maintaining the SDSR offers a case study for academic medical centers developing data science and research informatics infrastructure.

Although the clinical application and transformation of exosomes are still in the exploration stage, the prospects are promising and have a profound impact on the future transformation medicine of exosomes. However, due to the limitation of production and poor targeting ability of exosomes, the extensive and rich biological functions of exosomes are restricted, and the potential of clinical transformation is limited. The current research is committed to solving the above problems and expanding the clinical application value, but it lacks an extensive, multi-angle, and comprehensive systematic summary and prospect. Therefore, we reviewed the current optimization strategies of exosomes in medical applications, including the exogenous treatment of parent cells and the improvement of extraction methods, and compared their advantages and disadvantages. Subsequently, the targeting ability was improved by carrying drugs and engineering the structure of exosomes to solve the problem of poor targeting ability in clinical transformation. In addition, we discussed other problems that may exist in the application of exosomes. Although the clinical application and transformation of exosomes are still in the exploratory stage, the prospects are promising and have a profound impact on drug delivery, clinical diagnosis and treatment, and regenerative medicine.

Traditional pathology approaches have played an integral role in the delivery of diagnosis, semi-quantitative or qualitative assessment of protein expression, and classification of disease. Technological advances and the increased focus on precision medicine have recently paved the way for the development of digital pathology-based approaches for quantitative pathologic assessments, namely whole slide imaging and artificial intelligence (AI)-based solutions, allowing us to explore and extract information beyond human visual perception. Within the field of immuno-oncology, the application of such methodologies in drug development and translational research have created invaluable opportunities for deciphering complex pathophysiology and the discovery of novel biomarkers and drug targets. With an increasing number of treatment options available for any given disease, practitioners face the growing challenge of selecting the most appropriate treatment for each patient. The ever-increasing utilization of AI-based approaches substantially expands our understanding of the tumor microenvironment, with digital approaches to patient stratification and selection for diagnostic assays supporting the identification of the optimal treatment regimen based on patient profiles. This review provides an overview of the opportunities and limitations around implementing AI-based methods in biomarker discovery and patient selection and discusses how advances in digital pathology and AI should be considered in the current landscape of translational medicine, touching on challenges this technology may face if adopted in clinical settings. The traditional role of pathologists in delivering accurate diagnoses or assessing biomarkers for companion diagnostics may be enhanced in precision, reproducibility, and scale by AI-powered analysis tools.

There is growing evidence to the importance of translational science and medicine in the improvement of patient outcome, even though the definitions of translational science, translational medicine, and clinical and translational medicine need to be further clarified. In the present perspective, we collected commentaries and descriptions about clinical and translational medicine from some members of Clinical and Translational Medicine editorial board to stimulate the discussion and help the understanding better.

Clinical and translational medicine is an ongoing opportunity and challenge to transition science from bench to patient, from patient to bench, and from patient to policy for scientists and clinicians. Of the many objectives, the most important one is to improve the quality of life in patients, including living duration, dignity, emotional well-being, psychiatric and physical health, as well as independency. Emphasis and redefining concepts has allowed clinical and translational medicine to experience exciting and productive developments. Further discoveries, innovations, validations, and developments in clinical and translational medicine are expected in 2020. The present editorial aims to present the top three most discussed topics with high expectations for translation from clinical investigations and trials into practice and application.

The fusion of insights from the comprehensive global burden of disease (GBD) study and the advanced artificial intelligence of open artificial intelligence (AI) chat generative pre-trained transformer version 4 (ChatGPT-4) brings the potential to transform personalized healthcare planning. By integrating the data-driven findings of the GBD study with the powerful conversational capabilities of ChatGPT-4, healthcare professionals can devise customized healthcare plans that are adapted to patients' lifestyles and preferences. We propose that this innovative partnership can lead to the creation of a novel AI-assisted personalized disease burden (AI-PDB) assessment and planning tool. For the successful implementation of this unconventional technology, it is crucial to ensure continuous and accurate updates, expert supervision, and address potential biases and limitations. Healthcare professionals and stakeholders should have a balanced and dynamic approach, emphasizing interdisciplinary collaborations, data accuracy, transparency, ethical compliance, and ongoing training. By investing in the unique strengths of both ChatGPT-4, especially its newly introduced features such as live internet browsing or plugins, and the GBD study, we may enhance personalized healthcare planning. This innovative approach has the potential to improve patient outcomes and optimize resource utilization, as well as pave the way for the worldwide implementation of precision medicine, thereby revolutionizing the existing healthcare landscape. However, to fully harness these benefits at both the global and individual levels, further research and development are warranted. This will ensure that we effectively tap into the potential of this synergy, bringing societies closer to a future where personalized healthcare is the norm rather than the exception.