Machine learning discriminates a movement disorder in a zebrafish model of Parkinson's disease

Machine learning enables high-throughput, low-replicate screening for novel anti-seizure targets and compounds using combined movement and calcium fluorescence in larval zebrafish

Identifying new anti-seizure medications (ASMs) is difficult due to limitations in animal-based assays. Zebrafish (Danio rerio) serve as a model for chemical and genetic seizures, but current methods for detecting anti-seizure responses are limited by incomplete detection of anti-seizure responses (locomotor assays) or low-throughput (electrophysiology, fluorescence microscopy). To overcome these challenges, we developed a novel high-throughput method using combined locomotor and calcium fluorescence data from unrestrained larval zebrafish in a 96-well plate reader. Custom software tracked fish movement and fluorescence changes (deltaF/F0) from high-speed time-series, and logistic classifiers trained with elastic net regression distinguished seizure-like activity in response to the GABAA receptor antagonist pentylenetetrazole (PTZ). A classifier using combined data ("PTZ M + F"; AUC-ROC: 0.98; F1: 0.912) outperformed movement-only ("PTZ M"; AUC-ROC: 0.9) and fluorescence-only classifiers ("PTZ F"; AUC-ROC 0.96). Seizure-like event rate increased in proportion to PTZ concentration, and was suppressed by valproic acid (VPA). Meanwhile, TGB selectively reduced events defined by the "PTZ M + F″ classifier, paralleling previous reports that TGB reduces electrographic but not locomotor seizures. Using bootstrap simulation, we calculated statistical power and demonstrated reliable detection of ASM effects with as few as N = 4 replicates. In a test screen, 4 out of 5 ASMs were detected. This high-throughput approach combines previously orthogonal assays for zebrafish ASM screening.

Drug repositioning for Parkinson's disease: An emphasis on artificial intelligence approaches

Parkinson's disease (PD) is one of the most incapacitating neurodegenerative diseases (NDDs). PD is the second most common NDD worldwide which affects approximately 1-2 percent of people over 65 years. It is an attractive pursuit for artificial intelligence (AI) to contribute to and evolve PD treatments through drug repositioning by repurposing existing drugs, shelved drugs, or even candidates that do not meet the criteria for clinical trials. A search was conducted in three databases Web of Science, Scopus, and PubMed. We reviewed the data related to the last years (1975-present) to identify those drugs currently being proposed for repositioning in PD. Moreover, we reviewed the present status of the computational approach, including AI/Machine Learning (AI/ML)-powered pharmaceutical discovery efforts and their implementation in PD treatment. It was found that the number of drug repositioning studies for PD has increased recently. Repositioning of drugs in PD is taking off, and scientific communities are increasingly interested in communicating its results and finding effective treatment alternatives for PD. A better chance of success in PD drug discovery has been made possible due to AI/ML algorithm advancements. In addition to the experimentation stage of drug discovery, it is also important to leverage AI in the planning stage of clinical trials to make them more effective. New AI-based models or solutions that increase the success rate of drug development are greatly needed.

Anxiety in aquatics: Leveraging machine learning models to predict adult zebrafish behavior

Accurate analysis of anxiety behaviors in animal models is pivotal for advancing neuroscience research and drug discovery. This study compares the potential of DeepLabCut, ZebraLab, and machine learning models to analyze anxiety-related behaviors in adult zebrafish. Using a dataset comprising video recordings of unstressed and pre-stressed zebrafish, we extracted features such as total inactivity duration/immobility, time spent at the bottom, time spent at the top and turn angles (large and small). We observed that the data obtained using DeepLabCut and ZebraLab were highly correlated. Using this data, we annotated behaviors as anxious and not anxious and trained several machine learning models, including Logistic Regression, Decision Tree, K-Nearest Neighbours (KNN), Random Forests, Naive Bayes Classifiers, and Support Vector Machines (SVMs). The effectiveness of these machine learning models was validated and tested on independent datasets. We found that some machine learning models, such as Decision Tree and Random Forests, performed excellently to differentiate between anxious and non-anxious behavior, even in the control group, where the differences between subjects were more subtle. Our findings show that upcoming technologies, such as machine learning models, are able to effectively and accurately analyze anxiety behaviors in zebrafish and provide a cost-effective method to analyze animal behavior.

Accelerated Sarcopenia Phenotype in the DJ-1/Park7-Knockout Zebrafish

Age-dependent loss of muscle mass and function is associated with oxidative stress. DJ-1/Park7 acts as an antioxidant through multiple signalling pathways. DJ-1-knockout zebrafish show a decline in swimming performance and loss of weight gain between 6 and 9 months of age. Here, we address the degree to which this is associated with muscle degeneration and identify molecular changes preceding dysregulation of muscle performance. Loss of DJ-1 reduced the skeletal muscle fibre cross-section area. The highly mitochondrial-dependent red slow muscle was more affected than the white muscle, and degeneration of sub-sarcolemma red muscle mitochondria was observed. Using TandemMassTag-based quantitative proteomics, we identified a total of 3721 proteins in the multiplex sample of 4 and 12-month-old muscles. A total of 68 proteins, mainly associated with inflammation and mitochondrial function, were dysregulated in the young DJ-1-null adults, with Annexin A3, Sphingomyelin phosphodiesterase acid-like 3B, Complement C3a, and 2,4-dienoyl CoA reductase 1 being the most affected. The loss of DJ-1 also accelerated molecular features associated with sarcopenia, such as a decrease in the NAD+/NADH ratio and a reduction in Prostaglandin reductase 2 and Cytosolic glycerol-3-phosphate dehydrogenase levels. In view of the experimental power of zebrafish, the DJ-1-null zebrafish makes a valuable model for understanding the connection between oxidative stress and age-dependent muscle loss and function.

Deep learning dives: Predicting anxiety in zebrafish through novel tank assay analysis

Behavior is fundamental to neuroscience research, providing insights into the mechanisms underlying thoughts, actions and responses. Various model organisms, including mice, flies, and fish, are employed to understand these mechanisms. Zebrafish, in particular, serve as a valuable model for studying anxiety-like behavior, typically measured through the novel tank diving (NTD) assay. Traditional methods for analyzing NTD assays are either manually intensive or costly when using specialized software. To address these limitations, it is useful to develop methods for the automated analysis of zebrafish NTD assays using deep-learning models. In this study, we classified zebrafish based on their anxiety levels using DeepLabCut. Subsequently, based on a training dataset of image frames, we compared deep-learning models to identify the model best suited to classify zebrafish as anxious or non anxious and found that specific architectures, such as InceptionV3, are able to effectively perform this classification task. Our findings suggest that these deep learning models hold promise for automated behavioral analysis in zebrafish, offering an efficient and cost-effective alternative to traditional methods.

Deep phenotypic profiling of neuroactive drugs in larval zebrafish

Behavioral larval zebrafish screens leverage a high-throughput small molecule discovery format to find neuroactive molecules relevant to mammalian physiology. We screen a library of 650 central nervous system active compounds in high replicate to train deep metric learning models on zebrafish behavioral profiles. The machine learning initially exploited subtle artifacts in the phenotypic screen, necessitating a complete experimental re-run with rigorous physical well-wise randomization. These large matched phenotypic screening datasets (initial and well-randomized) provide a unique opportunity to quantify and understand shortcut learning in a full-scale, real-world drug discovery dataset. The final deep metric learning model substantially outperforms correlation distance-the canonical way of computing distances between profiles-and generalizes to an orthogonal dataset of diverse drug-like compounds. We validate predictions by prospective in vitro radio-ligand binding assays against human protein targets, achieving a hit rate of 58% despite crossing species and chemical scaffold boundaries. These neuroactive compounds exhibit diverse chemical scaffolds, demonstrating that zebrafish phenotypic screens combined with metric learning achieve robust scaffold hopping capabilities.

Application of machine learning in the study of development, behavior, nerve, and genotoxicity of zebrafish

Machine learning (ML) as a novel model-based approach has been used in studying aquatic toxicology in the environmental field. Zebrafish, as an ideal model organism in aquatic toxicology research, has been widely used to study the toxic effects of various pollutants. However, toxicity testing on organisms may cause significant harm, consume considerable time and resources, and raise ethical concerns. Therefore, ML is used in related research to reduce animal experiments and assist researchers in conducting toxicological research. Although ML techniques have matured in various fields, research on ML-based aquatic toxicology is still in its infancy due to the lack of comprehensive large-scale toxicity databases for environmental pollutants and model organisms. Therefore, to better understand the recent research progress of ML in studying the development, behavior, nerve, and genotoxicity of zebrafish, this review mainly focuses on using ML modeling to assess and predict the toxic effects of zebrafish exposure to different toxic chemicals. Meanwhile, the opportunities and challenges faced by ML in the field of toxicology were analyzed. Finally, suggestions and perspectives were proposed for the toxicity studies of ML on zebrafish in future applications.

Advances in animal models of Parkinson's disease

Parkinson's disease is a complex neurodegenerative disease characterized by progressive movement impairments. Predominant symptoms encompass resting tremor, bradykinesia, limb rigidity, and postural instability. In addition, it also includes a series of non-motor symptoms such as sleep disorders, hyposmia, gastrointestinal dysfunction, autonomic dysfunction and cognitive impairment. Pathologically, the disease manifests through dopaminergic neuronal loss and the presence of Lewy bodies. At present, no significant breakthrough has been achieved in clinical Parkinson's disease treatment. Exploring treatment modalities necessitate the establishment of scientifically sound animal models. In recent years, researchers have focused on replicating the symptoms of human Parkinson's disease, resulting in the establishment of various experimental animal models primarily through drugs and transgenic methods to mimic relevant pathologies and identify more effective treatments. This review examines traditional neurotoxin and transgenic animal models as well as α-synuclein pre-formed fibrils models, non-human primate models and non-mammalian specie models. Additionally, it introduces emerging models, including models based on optogenetics, induced pluripotent stem cells, and gene editing, aiming to provide a reference for the utilization of experimental animal models and clinical research for researchers in this field.

Nicotinamide riboside ameliorates survival time and motor dysfunction in an MPTP-Induced Parkinson's disease zebrafish model through effects on glucose metabolism and endoplasmic reticulum stress

Nicotinamide riboside (NR) is a precursor and exogenous supplement of nicotinamide adenine dinucleotide (NAD+). NR has been shown to play a beneficial role in a variety of neurodegenerative diseases. A phase 1 clinical trial identified NR as a potential neuroprotective therapy for Parkinson's disease (PD). However, the mechanism of action of NR in PD has not been fully elucidated. Therefore, the present study aimed to investigate the potential effects of NR on a 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced PD model in zebrafish and its underlying mechanisms. The results showed that NR improved motor dysfunction, survival time, dopamine neurons, and peripheral neurons, as well as the NAD+ levels in the MPTP-affected PD zebrafish model. In addition, transcriptome sequencing analysis revealed that, after NR treatment, differentially expressed genes were significantly enriched in the glucose metabolism and protein processing pathways in the endoplasmic reticulum (ER). Quantitative PCR (qPCR) revealed that the mRNA levels of the glycoheterotrophic enzyme (involved in glucose metabolism) were significantly decreased, and the glycolytic enzyme mRNA expression levels were significantly increased. The results of the non-targeted metabolomic analysis showed that NR treatment significantly increased the levels of metabolites such as nicotinic acid ,nicotinamide, d-glucose (from the gluconeogenesis and glycolysis metabolism pathways) and some glucogenic amino acids, such as glutamine. Importantly, NR ameliorated MPTP-induced endoplasmic reticulum stress (ERS) in the PD zebrafish model through the Perk-Eif2α-Atf4-Chop pathway. These results highlight the neuroprotective effect of NR in the present PD zebrafish model through modulation of glucose metabolism and ERS via the Perk-Eif2α-Atf4-Chop pathway and provide valuable mechanistic insights into the treatment of PD.

Three dimensional and four dimensional live imaging to study mechanisms of progressive neurodegeneration

Neurodegenerative diseases are complex and progressive, posing challenges to their study and understanding. Recent advances in microscopy imaging technologies have enabled the exploration of neurons in three spatial dimensions (3D) over time (4D). When applied to 3D cultures, tissues, or animals, these technologies can provide valuable insights into the dynamic and spatial nature of neurodegenerative diseases. This review focuses on the use of imaging techniques and neurodegenerative disease models to study neurodegeneration in 4D. Imaging techniques such as confocal microscopy, two-photon microscopy, miniscope imaging, light sheet microscopy, and robotic microscopy offer powerful tools to visualize and analyze neuronal changes over time in 3D tissue. Application of these technologies to in vitro models of neurodegeneration such as mouse organotypic culture systems and human organoid models provide versatile platforms to study neurodegeneration in a physiologically relevant context. Additionally, use of 4D imaging in vivo, including in mouse and zebrafish models of neurodegenerative diseases, allows for the investigation of early dysfunction and behavioral changes associated with neurodegeneration. We propose that these studies have the power to overcome the limitations of two-dimensional monolayer neuronal cultures and pave the way for improved understanding of the dynamics of neurodegenerative diseases and the development of effective therapeutic strategies.

Towards improved screening of toxins for Parkinson's risk

Parkinson's disease (PD) is a chronic, progressive and disabling neurodegenerative disorder. The prevalence of PD has risen considerably over the past decades. A growing body of evidence suggest that exposure to environmental toxins, including pesticides, solvents and heavy metals (collectively called toxins), is at least in part responsible for this rapid growth. It is worrying that the current screening procedures being applied internationally to test for possible neurotoxicity of specific compounds offer inadequate insights into the risk of developing PD in humans. Improved screening procedures are therefore urgently needed. Our review first substantiates current evidence on the relation between exposure to environmental toxins and the risk of developing PD. We subsequently propose to replace the current standard toxin screening by a well-controlled multi-tier toxin screening involving the following steps: in silico studies (tier 1) followed by in vitro tests (tier 2), aiming to prioritize agents with human relevant routes of exposure. More in depth studies can be undertaken in tier 3, with whole-organism (in)vertebrate models. Tier 4 has a dedicated focus on cell loss in the substantia nigra and on the presumed mechanisms of neurotoxicity in rodent models, which are required to confirm or refute the possible neurotoxicity of any individual compound. This improved screening procedure should not only evaluate new pesticides that seek access to the market, but also critically assess all pesticides that are being used today, acknowledging that none of these has ever been proven to be safe from a perspective of PD. Importantly, the improved screening procedures should not just assess the neurotoxic risk of isolated compounds, but should also specifically look at the cumulative risk conveyed by exposure to commonly used combinations of pesticides (cocktails). The worldwide implementation of such an improved screening procedure, would be an essential step for policy makers and governments to recognize PD-related environmental risk factors.

Unlocking the Potential of Zebrafish Research with Artificial Intelligence: Advancements in Tracking, Processing, and Visualization

Zebrafish have become a widely accepted model organism for biomedical research due to their strong cortisol stress response, behavioral strain differences, and sensitivity to both drug treatments and predators. However, experimental zebrafish studies generate substantial data that must be analyzed through objective, accurate, and repeatable analysis methods. Recently, advancements in artificial intelligence (AI) have enabled automated tracking, image recognition, and data analysis, leading to more efficient and insightful investigations. In this review, we examine key AI applications in zebrafish research, including behavior analysis, genomics, and neuroscience. With the development of deep learning technology, AI algorithms have been used to precisely analyze and identify images of zebrafish, enabling automated testing and analysis. By applying AI algorithms in genomics research, researchers have elucidated the relationship between genes and biology, providing a better basis for the development of disease treatments and gene therapies. Additionally, the development of more effective neuroscience tools could help researchers better understand the complex neural networks in the zebrafish brain. In the future, further advancements in AI technology are expected to enable more extensive and in-depth medical research applications in zebrafish, improving our understanding of this important animal model. This review highlights the potential of AI technology in achieving the full potential of zebrafish research by enabling researchers to efficiently track, process, and visualize the outcomes of their experiments.

Myofibrillar myopathy hallmarks associated with ZAK deficiency

The ZAK gene encodes two functionally distinct kinases, ZAKα and ZAKβ. Homozygous loss of function mutations affecting both isoforms causes a congenital muscle disease. ZAKβ is the only isoform expressed in skeletal muscle and is activated by muscle contraction and cellular compression. The ZAKβ substrates in skeletal muscle or the mechanism whereby ZAKβ senses mechanical stress remains to be determined. To gain insights into the pathogenic mechanism, we exploited ZAK-deficient cell lines, zebrafish, mice and a human biopsy. ZAK-deficient mice and zebrafish show a mild phenotype. In mice, comparative histopathology data from regeneration, overloading, ageing and sex conditions indicate that while age and activity are drivers of the pathology, ZAKβ appears to have a marginal role in myoblast fusion in vitro or muscle regeneration in vivo. The presence of SYNPO2, BAG3 and Filamin C (FLNC) in a phosphoproteomics assay and extended analyses suggested a role for ZAKβ in the turnover of FLNC. Immunofluorescence analysis of muscle sections from mice and a human biopsy showed evidence of FLNC and BAG3 accumulations as well as other myofibrillar myopathy markers. Moreover, endogenous overloading of skeletal muscle exacerbated the presence of fibres with FLNC accumulations in mice, indicating that ZAKβ signalling is necessary for an adaptive turnover of FLNC that allows for the normal physiological response to sustained mechanical stress. We suggest that accumulation of mislocalized FLNC and BAG3 in highly immunoreactive fibres contributes to the pathogenic mechanism of ZAK deficiency.

DANGER analysis: risk-averse on/off-target assessment for CRISPR editing without a reference genome

Motivation

The CRISPR-Cas9 system has successfully achieved site-specific gene editing in organisms ranging from humans to bacteria. The technology efficiently generates mutants, allowing for phenotypic analysis of the on-target gene. However, some conventional studies did not investigate whether deleterious off-target effects partially affect the phenotype.

Results

Herein, we present a novel phenotypic assessment of CRISPR-mediated gene editing: Deleterious and ANticipatable Guides Evaluated by RNA-sequencing (DANGER) analysis. Using RNA-seq data, this bioinformatics pipeline can elucidate genomic on/off-target sites on mRNA-transcribed regions related to expression changes and then quantify phenotypic risk at the gene ontology term level. We demonstrated the risk-averse on/off-target assessment in RNA-seq data from gene-edited samples of human cells and zebrafish brains. Our DANGER analysis successfully detected off-target sites, and it quantitatively evaluated the potential contribution of deleterious off-targets to the transcriptome phenotypes of the edited mutants. Notably, DANGER analysis harnessed de novo transcriptome assembly to perform risk-averse on/off-target assessments without a reference genome. Thus, our resources would help assess genome editing in non-model organisms, individual human genomes, and atypical genomes from diseases and viruses. In conclusion, DANGER analysis facilitates the safer design of genome editing in all organisms with a transcriptome.

Availability and implementation

The Script for the DANGER analysis pipeline is available at https://github.com/KazukiNakamae/DANGER_analysis. In addition, the software provides a tutorial on reproducing the results presented in this article on the Readme page. The Docker image of DANGER_analysis is also available at https://hub.docker.com/repository/docker/kazukinakamae/dangeranalysis/general.

Development of CRISPR Cas9, spin-off technologies and their application in model construction and potential therapeutic methods of Parkinson's disease

Parkinson's disease (PD) is one of the most common degenerative diseases. It is most typically characterized by neuronal death following the accumulation of Lewis inclusions in dopaminergic neurons in the substantia nigra region, with clinical symptoms such as motor retardation, autonomic dysfunction, and dystonia spasms. The exact molecular mechanism of its pathogenesis has not been revealed up to now. And there is a lack of effective treatments for PD, which places a burden on patients, families, and society. CRISPR Cas9 is a powerful technology to modify target genomic sequence with rapid development. More and more scientists utilized this technique to perform research associated neurodegenerative disease including PD. However, the complexity involved makes it urgent to organize and summarize the existing findings to facilitate a clearer understanding. In this review, we described the development of CRISPR Cas9 technology and the latest spin-off gene editing systems. Then we focused on the application of CRISPR Cas9 technology in PD research, summarizing the construction of the novel PD-related medical models including cellular models, small animal models, large mammal models. We also discussed new directions and target molecules related to the use of CRISPR Cas9 for PD treatment from the above models. Finally, we proposed the view about the directions for the development and optimization of the CRISPR Cas9 technology system, and its application to PD and gene therapy in the future. All these results provided a valuable reference and enhanced in understanding for studying PD.

Fish Models for Exploring Mitochondrial Dysfunction Affecting Neurodegenerative Disorders

Neurodegenerative disorders are characterized by the progressive loss of neuronal structure or function, resulting in memory loss and movement disorders. Although the detailed pathogenic mechanism has not been elucidated, it is thought to be related to the loss of mitochondrial function in the process of aging. Animal models that mimic the pathology of a disease are essential for understanding human diseases. In recent years, small fish have become ideal vertebrate models for human disease due to their high genetic and histological homology to humans, ease of in vivo imaging, and ease of genetic manipulation. In this review, we first outline the impact of mitochondrial dysfunction on the progression of neurodegenerative diseases. Then, we highlight the advantages of small fish as model organisms, and present examples of previous studies regarding mitochondria-related neuronal disorders. Lastly, we discuss the applicability of the turquoise killifish, a unique model for aging research, as a model for neurodegenerative diseases. Small fish models are expected to advance our understanding of the mitochondrial function in vivo, the pathogenesis of neurodegenerative diseases, and be important tools for developing therapies to treat diseases.

The Importance of Animal Models in Biomedical Research: Current Insights and Applications

Animal research is considered a key element in advance of biomedical science. Although its use is controversial and raises ethical challenges, the contribution of animal models in medicine is essential for understanding the physiopathology and novel treatment alternatives for several animal and human diseases. Current pandemics’ pathology, such as the 2019 Coronavirus disease, has been studied in primate, rodent, and porcine models to recognize infection routes and develop therapeutic protocols. Worldwide issues such as diabetes, obesity, neurological disorders, pain, rehabilitation medicine, and surgical techniques require studying the process in different animal species before testing them on humans. Due to their relevance, this article aims to discuss the importance of animal models in diverse lines of biomedical research by analyzing the contributions of the various species utilized in science over the past five years about key topics concerning human and animal health.

The impact of stress and anesthesia on animal models of infectious disease

Stress and general anesthesia have an impact on the functional response of the organism due to the detrimental effects on cardiovascular, immunological, and metabolic function, which could limit the organism's response to an infectious event. Animal studies have formed an essential step in understanding and mitigating infectious diseases, as the complexities of physiology and immunity cannot yet be replicated in vivo. Using animals in research continues to come under increasing societal scrutiny, and it is therefore crucial that the welfare of animals used in disease research is optimized to meet both societal expectations and improve scientific outcomes. Everyday management and procedures in animal studies are known to cause stress, which can not only cause poorer welfare outcomes, but also introduces variables in disease studies. Whilst general anesthesia is necessary at times to reduce stress and enhance animal welfare in disease research, evidence of physiological and immunological disruption caused by general anesthesia is increasing. To better understand and quantify the effects of stress and anesthesia on disease study and welfare outcomes, utilizing the most appropriate animal monitoring strategies is imperative. This article aims to analyze recent scientific evidence about the impact of stress and anesthesia as uncontrolled variables, as well as reviewing monitoring strategies and technologies in animal models during infectious diseases.

Advanced Techniques for the Intelligent Diagnosis of Fish Diseases: A Review

Aquatic products, as essential sources of protein, have attracted considerable concern by producers and consumers. Precise fish disease prevention and treatment may provide not only healthy fish protein but also ecological and economic benefits. However, unlike intelligent two-dimensional diagnoses of plants and crops, one of the most serious challenges confronted in intelligent aquaculture diagnosis is its three-dimensional space. Expert systems have been applied to diagnose fish diseases in recent decades, allowing for restricted diagnosis of certain aquaculture. However, this method needs aquaculture professionals and specialists. In addition, diagnosis speed and efficiency are limited. Therefore, developing a new quick, automatic, and real-time diagnosis approach is very critical. The integration of image-processing and computer vision technology intelligently allows the diagnosis of fish diseases. This study comprehensively reviews image-processing technology and image-based fish disease detection methods, and analyzes the benefits and drawbacks of each diagnostic approach in different environments. Although it is widely acknowledged that there are many approaches for disease diagnosis and pathogen identification, some improvements in detection accuracy and speed are still needed. Constructing AR 3D images of fish diseases, standard and shared datasets, deep learning, and data fusion techniques will be helpful in improving the accuracy and speed of fish disease diagnosis.

Fish Behavior as a Neural Proxy to Reveal Physiological States

Behaviors are the integrative outcomes of the nervous system, which senses and responds to the internal physiological status and external stimuli. Teleosts are aquatic organisms which are more easily affected by the surrounding environment compared to terrestrial animals. To date, behavioral tests have been widely used to assess potential environmental risks using fish as model animals. In this review, we summarized recent studies regarding the effects of internal and external stimuli on fish behaviors. We concluded that behaviors reflect environmental and physiological changes, which have possible implications for environmental and physiological assessments.

Zebrafish and Medaka: Important Animal Models for Human Neurodegenerative Diseases

Animal models of human neurodegenerative disease have been investigated for several decades. In recent years, zebrafish (Danio rerio) and medaka (Oryzias latipes) have become popular in pathogenic and therapeutic studies about human neurodegenerative diseases due to their small size, the optical clarity of embryos, their fast development, and their suitability to large-scale therapeutic screening. Following the emergence of a new generation of molecular biological technologies such as reverse and forward genetics, morpholino, transgenesis, and gene knockout, many human neurodegenerative disease models, such as Parkinson's, Huntington's, and Alzheimer's, were constructed in zebrafish and medaka. These studies proved that zebrafish and medaka genes are functionally conserved in relation to their human homologues, so they exhibit similar neurodegenerative phenotypes to human beings. Therefore, fish are a suitable model for the investigation of pathologic mechanisms of neurodegenerative diseases and for the large-scale screening of drugs for potential therapy. In this review, we summarize the studies in modelling human neurodegenerative diseases in zebrafish and medaka in recent years.

Advances of Zebrafish in Neurodegenerative Disease: From Models to Drug Discovery

Neurodegenerative disease (NDD), including Alzheimer’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis, are characterized by the progressive loss of neurons which leads to the decline of motor and/or cognitive function. Currently, the prevalence of NDD is rapidly increasing in the aging population. However, valid drugs or treatment for NDD are still lacking. The clinical heterogeneity and complex pathogenesis of NDD pose a great challenge for the development of disease-modifying therapies. Numerous animal models have been generated to mimic the pathological conditions of these diseases for drug discovery. Among them, zebrafish (Danio rerio) models are progressively emerging and becoming a powerful tool for in vivo study of NDD. Extensive use of zebrafish in pharmacology research or drug screening is due to the high conserved evolution and 87% homology to humans. In this review, we summarize the zebrafish models used in NDD studies, and highlight the recent findings on pharmacological targets for NDD treatment. As high-throughput platforms in zebrafish research have rapidly developed in recent years, we also discuss the application prospects of these new technologies in future NDD research.

Animal Models of Autosomal Recessive Parkinsonism

Parkinson’s disease (PD) is the most common neurodegenerative movement disorder. The neuropathological hallmark of the disease is the loss of dopamine neurons of the substantia nigra pars compacta. The clinical manifestations of PD are bradykinesia, rigidity, resting tremors and postural instability. PD patients often display non-motor symptoms such as depression, anxiety, weakness, sleep disturbances and cognitive disorders. Although, in 90% of cases, PD has a sporadic onset of unknown etiology, highly penetrant rare genetic mutations in many genes have been linked with typical familial PD. Understanding the mechanisms behind the DA neuron death in these Mendelian forms may help to illuminate the pathogenesis of DA neuron degeneration in the more common forms of PD. A key step in the identification of the molecular pathways underlying DA neuron death, and in the development of therapeutic strategies, is the creation and characterization of animal models that faithfully recapitulate the human disease. In this review, we outline the current status of PD modeling using mouse, rat and non-mammalian models, focusing on animal models for autosomal recessive PD.

Zebrafish behavior feature recognition using three-dimensional tracking and machine learning

In this work, we aim to construct a new behavior analysis method by using machine learning. We used two cameras to capture three-dimensional (3D) tracking data of zebrafish, which were analyzed using fuzzy adaptive resonance theory (FuzzyART), a type of machine learning algorithm, to identify specific behavioral features. The method was tested based on an experiment in which electric shocks were delivered to zebrafish and zebrafish swimming was tracked in 3D simultaneously to find electric shock-associated behaviors. By processing the obtained data with FuzzyART, we discovered that distinguishing behaviors were statistically linked to the electric shock based on the machine learning algorithm. Moreover, our system could accept user-supplied data for detection and quantitative analysis of the behavior features, such as the behavior features defined by the 3D tracking analysis above. This system could be applied to discover new distinct behavior features in mutant zebrafish and used for drug administration screening and cognitive ability tests of zebrafish in the future.

Loss of park7 activity has differential effects on expression of iron responsive element (IRE) gene sets in the brain transcriptome in a zebrafish model of Parkinson's disease

Mutation of the gene PARK7 (DJ1) causes monogenic autosomal recessive Parkinson's disease (PD) in humans. Subsequent alterations of PARK7 protein function lead to mitochondrial dysfunction, a major element in PD pathology. Homozygous mutants for the PARK7-orthologous genes in zebrafish, park7, show changes to gene expression in the oxidative phosphorylation pathway, supporting that disruption of energy production is a key feature of neurodegeneration in PD. Iron is critical for normal mitochondrial function, and we have previously used bioinformatic analysis of IRE-bearing transcripts in brain transcriptomes to find evidence supporting the existence of iron dyshomeostasis in Alzheimer's disease. Here, we analysed IRE-bearing transcripts in the transcriptome data from homozygous park7-/- mutant zebrafish brains. We found that the set of genes with "high quality" IREs in their 5' untranslated regions (UTRs, the HQ5'IRE gene set) was significantly altered in these 4-month-old park7-/- brains. However, sets of genes with IREs in their 3' UTRs appeared unaffected. The effects on HQ5'IRE genes are possibly driven by iron dyshomeostasis and/or oxidative stress, but illuminate the existence of currently unknown mechanisms with differential overall effects on 5' and 3' IREs.

Artificial intelligence and machine learning-aided drug discovery in central nervous system diseases: State-of-the-arts and future directions

Neurological disorders significantly outnumber diseases in other therapeutic areas. However, developing drugs for central nervous system (CNS) disorders remains the most challenging area in drug discovery, accompanied with the long timelines and high attrition rates. With the rapid growth of biomedical data enabled by advanced experimental technologies, artificial intelligence (AI) and machine learning (ML) have emerged as an indispensable tool to draw meaningful insights and improve decision making in drug discovery. Thanks to the advancements in AI and ML algorithms, now the AI/ML-driven solutions have an unprecedented potential to accelerate the process of CNS drug discovery with better success rate. In this review, we comprehensively summarize AI/ML-powered pharmaceutical discovery efforts and their implementations in the CNS area. After introducing the AI/ML models as well as the conceptualization and data preparation, we outline the applications of AI/ML technologies to several key procedures in drug discovery, including target identification, compound screening, hit/lead generation and optimization, drug response and synergy prediction, de novo drug design, and drug repurposing. We review the current state-of-the-art of AI/ML-guided CNS drug discovery, focusing on blood-brain barrier permeability prediction and implementation into therapeutic discovery for neurological diseases. Finally, we discuss the major challenges and limitations of current approaches and possible future directions that may provide resolutions to these difficulties.

The Promise of the Zebrafish Model for Parkinson's Disease: Today's Science and Tomorrow's Treatment

The second most prevalent neurodegenerative disorder in the elderly is Parkinson's disease (PD). Its etiology is unclear and there are no available disease-modifying medicines. Therefore, more evidence is required concerning its pathogenesis. The use of the neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) is the basis of most animal models of PD. MPTP is metabolized by monoamine oxidase B (MAO B) to MPP + and induces the loss of dopaminergic neurons in the substantia nigra in mammals. Zebrafish have been commonly used in developmental biology as a model organism, but owing to its perfect mix of properties, it is now emerging as a model for human diseases. Zebrafish (Danio rerio) are cheap and easy to sustain, evolve rapidly, breed transparent embryos in large amounts, and are readily manipulated by different methods, particularly genetic ones. Furthermore, zebrafish are vertebrate species and mammalian findings obtained from zebrafish may be more applicable than those derived from genetic models of invertebrates such as Drosophila melanogaster and Caenorhabditis elegans. The resemblance cannot be taken for granted, however. The goal of the present review article is to highlight the promise of zebrafish as a PD animal model. As its aminergic structures, MPTP mode of action, and PINK1 roles mimic those of mammalians, zebrafish seems to be a viable model for studying PD. The roles of zebrafish MAO, however, vary from those of the two types of MAO present in mammals. The benefits unique to zebrafish, such as the ability to perform large-scale genetic or drug screens, should be exploited in future experiments utilizing zebrafish PD models.

First person – Gideon Hughes

First Person is a series of interviews with the first authors of a selection of papers published in Disease Models & Mechanisms, helping early-career researchers promote themselves alongside their papers. Gideon Hughes is first author on ‘Machine learning discriminates a movement disorder in a zebrafish model of Parkinson's disease’, published in DMM. Gideon conducted the research described in this article while a PhD student in Betsy Pownall's lab at the University of York, York, UK. He is now a postdoc in the lab of Henry Roehl at the University of Sheffield, Sheffield, UK, using the zebrafish as a model organism to study human disease and tissue regeneration, combining his research with his interest in computer science.