Deep Phenotyping Research Articles

Hereditary Ataxias in Argentina.

Hereditary or genetic ataxias are hundreds of disorders characterized by large phenotypic, genetic, and epidemiological heterogeneity. In Argentina, 35 genetic ataxias have been identified, with SCA1 (ATX-ATXN1), SCA2 (ATX-ATXN2), SCA3 (ATX-ATXN3), and Friedreich ataxia (ATX-FXN) as the most prevalent causes, reflecting the epidemiology of most Western European countries, the main origin of immigration to the country. Genetic diagnostic studies of ataxia cohorts in Argentina have found high rates of undiagnosed patients, ranging from 65 to 82%. Deep phenotyping, comprehensive genetic testing, and knowledge of the prevalence of different genetic ataxias are essential for an accurate diagnostic and treatment approach in clinical practice. This narrative review proposes a targeted, tiered genetic diagnostic approach for undiagnosed patients based on the Argentinian epidemiological and healthcare system data. Future national efforts should support comprehensive screening studies on ataxia cohorts, including testing for repeat expansions in RFC1 and FGF14 genes. In addition, establishing a trial-ready patient registry for genetic ataxias, enhancing networking with international clinical and research initiatives, and developing specialized centers for interdisciplinary care of genetic ataxia patients are recommended.

The RaDiCo information system for rare disease cohorts

BackgroundRare diseases (RDs) clinical care and research face several challenges. Patients are dispersed over large geographic areas, their number per disease is limited, just like the number of researchers involved. Current databases as well as biological collections, when existing, are generally local, of modest size, incomplete, of uneven quality, heterogeneous in format and content, and rarely accessible or standardised to support interoperability. Most disease phenotypes are complex corresponding to multi-systemic conditions, with insufficient interdisciplinary cooperation. Thus emerged the need to generate, within a coordinated, mutualised, secure and interoperable framework, high-quality data from national or international RD cohorts, based on deep phenotyping, including molecular analysis data, notably genotypic. The RaDiCo program objective was to create, under the umbrella of Inserm, a national operational platform dedicated to the development of RD e-cohorts. Its Information System (IS) is presented here.Material and methodsConstructed on the cloud computing principle, the RaDiCo platform was designed to promote mutualization and factorization of processes and services, for both clinical epidemiology support and IS. RaDiCo IS is based on an interoperability framework combining a unique RD identifier, data standardisation, FAIR principles, data exchange flows/processes and data security principles compliant with the European GDPR.ResultsRaDiCo IS favours a secure, open-source web application in order to implement and manage online databases and give patients themselves the opportunity to collect their data. It ensures a continuous monitoring of data quality and consistency over time. RaDiCo IS proved to be efficient, currently hosting 13 e-cohorts, covering 67 distinct RDs. As of April 2024, 8063 patients were recruited from 180 specialised RD sites spread across the national territory.DiscussionThe RaDiCo operational platform is equivalent to a national infrastructure. Its IS enables RD e-cohorts to be developed on a shared platform with no limit on size or number. Compliant with the GDPR, it is compatible with the French National Health Data Hub and can be extended to the RDs European Reference Networks (ERNs).ConclusionRaDiCo provides a robust IS, compatible with the French Data Hub and RDs ERNs, integrated on a RD platform that enables e-cohorts creation, monitoring and analysis.

Alignment of auditory artificial networks with massive individual fMRI brain data leads to generalisable improvements in brain encoding and downstream tasks

Abstract Artificial neural networks trained in the field of artificial intelligence (AI) have emerged as key tools to model brain processes, sparking the idea of aligning network representations with brain dynamics to enhance performance on AI tasks. While this concept has gained support in the visual domain, we investigate here the feasibility of creating auditory artificial neural models directly aligned with individual brain activity. This objective raises major computational challenges, as models have to be trained directly with brain data, which is typically collected at a much smaller scale than data used to train AI models. We aimed to answer two key questions: (1) Can brain alignment of auditory models lead to improved brain encoding for novel, previously unseen stimuli? (2) Can brain alignment lead to generalisable representations of auditory signals that are useful for solving a variety of complex auditory tasks? To answer these questions, we relied on two massive datasets: a deep phenotyping dataset from the Courtois neuronal modelling project, where six subjects watched four seasons (36 h) of the Friends TV series in functional magnetic resonance imaging and the HEAR benchmark, a large battery of downstream auditory tasks. We fine-tuned SoundNet, a small pretrained convolutional neural network with ~2.5 M parameters. Aligning SoundNet with brain data from three seasons of Friends led to substantial improvement in brain encoding in the fourth season, extending beyond auditory and visual cortices. We also observed consistent performance gains on the HEAR benchmark, particularly for tasks with limited training data, where brain-aligned models performed comparably with the best-performing models regardless of size. We finally compared individual and group models, finding that individual models often matched or outperformed group models in both brain encoding and downstream task performance, highlighting the data efficiency of fine-tuning with individual brain data. Our results demonstrate the feasibility of aligning artificial neural network representations with individual brain activity during auditory processing, and suggest that this alignment is particularly beneficial for tasks with limited training data. Future research is needed to establish whether larger models can achieve even better performance and whether the observed gains extend to other tasks, particularly in the context of few-shot learning.

Aarskog Syndrome: Deep Phenotyping and Genomic Landscape of a New Cohort Including Adult Patients.

Aarskog-Scott syndrome (AAS, MIM#305400) is an X-linked disorder characterized by recognizable facial features, short stature, and genitourinary and skeletal malformations. AAS is attributed to pathogenic variants in FGD1, and ~60 patients with a genetic diagnosis have been reported to date. We hereby present a molecularly confirmed cohort of 14 male AAS patients from 13 families. Among 14 patients, 12 were referred during childhood, while two were referred at adulthood due to infertility. Six out of 11 patients with available records had antenatal manifestations, comprising shortened tubular bones, growth restriction, polyhydramnios, pes equinovarus, increased nuchal translucency, fetal hypokinesia, echogenic intracardiac focus, and ambiguous genitalia. In addition to well-described AAS findings, distinctive features observed in multiple patients included variable skin findings (n = 5), renal malformations (n = 2), muscular build (n = 2), and infertility (n = 2). Cardiac (n = 4) and ocular manifestations (n = 6) were identified at significantly higher rates than previously reported. This cohort also presents new patients with osteochondritis dissecans and oligo/azoospermia, providing further evidence to acknowledge these once-reported findings as part of the disease spectrum. Eleven different FGD1 variants, including seven novel ones, were identified through targeted FGD1 sequencing. Two variants were found to be recurrent, detected in two independent families. Our study provides additional insights into the clinical and genotypic landscape of AAS through the largest molecularly confirmed cohort, including two adult patients.

Integration of multi-omics data and deep phenotyping provides insights into responses to single and combined abiotic stress in potato.

Potato (Solanum tuberosum) is highly water and space efficient but susceptible to abiotic stresses such as heat, drought, and flooding, which are severely exacerbated by climate change. Our understanding of crop acclimation to abiotic stress, however, remains limited. Here, we present a comprehensive molecular and physiological high-throughput profiling of potato (Solanum tuberosum, cv. Désirée) under heat, drought, and waterlogging applied as single stresses or in combinations designed to mimic realistic future scenarios. Stress responses were monitored via daily phenotyping and multi-omics analyses of leaf samples comprising proteomics, targeted transcriptomics, metabolomics, and hormonomics at several timepoints during and after stress treatments. Additionally, critical metabolites of tuber samples were analyzed at the end of the stress period. We performed integrative multi-omics data analysis using a bioinformatic pipeline that we established based on machine learning and knowledge networks. Waterlogging produced the most immediate and dramatic effects on potato plants, interestingly activating ABA responses similar to drought stress. In addition, we observed distinct stress signatures at multiple molecular levels in response to heat or drought and to a combination of both. In response to all treatments, we found a downregulation of photosynthesis at different molecular levels, an accumulation of minor amino acids, and diverse stress-induced hormones. Our integrative multi-omics analysis provides global insights into plant stress responses, facilitating improved breeding strategies toward climate-adapted potato varieties.

Deep phenotyping to understand hearing and hearing disorders: Protocol for a feasibility study.

Globally, hearing loss affects around 1.5 billion people, while tinnitus is estimated to impact around 740 million. More research is urgently needed to address the challenges presented by hearing loss, tinnitus, and other hearing-related conditions. Our plans for a Nottingham Hearing BioResource, providing research-willing volunteers and comprehensive tests of hearing and ear health repeated over time, has the potential to accelerate the field. The protocol described here is a feasibility study for this BioResource, specifically addressing questions of recruitment from the general population (i.e., outside of clinical audiology services or pathways). Participants with or without known hearing problems will be recruited for data collection. This study will quantify how feasible it will be to recruit and retain a large sample of the general population, and will suggest the demographic, and hearing condition status, distributions we could achieve for the BioResource. Data collection will involve a health and lifestyle questionnaire; cognitive assessment; five questionnaires about hearing loss, tinnitus, and hyperacusis; an estimation of lifetime noise exposure; a suite of in-depth audiological tests; and taking a hair sample. The same measurements will be taken on two separate occasions in person, and a third set of overlapping measurements will be taken remotely. Repeating the data collection will allow us to evaluate participant retention rates and establish the reliability of the measures. The findings from this feasibility study will allow us to assess which channels work well to recruit a diverse pool of participants, which, when used in conjunction with recruitment from clinic, will provide the basis for a recruitment strategy for our BioResource. In addition, we will gain useful insight into whether specific tests or questionnaires used in the feasibility study are suitable for inclusion in a deep phenotyping protocol.

Deep developmental phenotyping in children with tuberous sclerosis complex, with and without autism.

https://onlinelibrary.wiley.com/doi/10.1111/dmcn.16293

Deep Phenotyping of the Broader Autism Phenotype in Epilepsy: A Transdiagnostic Marker of Epilepsy and Autism Spectrum Disorder

Deep Phenotyping of the Broader Autism Phenotype in Epilepsy: A Transdiagnostic Marker of Epilepsy and Autism Spectrum Disorder

A high precision method of segmenting complex postures in Caenorhabditis elegans and deep phenotyping to analyze lifespan

In-depth exploration of the effects of genes on the development, physiology, and behavior of organisms requires high-precision phenotypic analysis. However, the overlap of body postures in group behavior and the similarity of movement patterns between strains pose challenges to accuracy analysis. To address this issue, we designed the WormYOLO model based on the YOLO architecture, which improves the segmentation performance of C .elegans and effectively handles overlapping poses in images. In detection and segmentation tasks, WormYOLO performs well on the more overlapping Mating dataset, with its object detection performance improving by 24.1% (mAP0.5:0.95) compared to Deep-worm-tracker, and its segmentation performance improving by 9.3% (mAP0.5:0.95) compared to WormSwin. In addition, we propose a more accurate novel bending counting algorithm. In experiments, WormYOLO segmented images, followed by a feature point extraction algorithm to identify changes in worm skeleton positions, ultimately quantifying behavioral features with a counting algorithm. We conducted analytical experiments on various mutant strains based on their motion characteristics, investigating behavioral differences among the strains and assessing the correlation between high-dimensional phenotypic traits and relative lifespan.

FlyVISTA, an integrated machine learning platform for deep phenotyping of sleep in Drosophila.

There is great interest in using genetically tractable organisms such as Drosophila to gain insights into the regulation and function of sleep. However, sleep phenotyping in Drosophila has largely relied on simple measures of locomotor inactivity. Here, we present FlyVISTA, a machine learning platform to perform deep phenotyping of sleep in flies. This platform comprises a high-resolution closed-loop video imaging system, coupled with a deep learning network to annotate 35 body parts, and a computational pipeline to extract behaviors from high-dimensional data. FlyVISTA reveals the distinct spatiotemporal dynamics of sleep and wake-associated microbehaviors at baseline, following administration of the sleep-inducing drug gaboxadol, and with dorsal fan-shaped body drivers. We identify a microbehavior ("haltere switch") exclusively seen during quiescence that indicates a deeper sleep stage. These results enable the rigorous analysis of sleep in Drosophila and set the stage for computational analyses of microbehaviors in quiescent animals.

Mapping variants in thyroid hormone transporter MCT8 to disease severity by genomic, phenotypic, functional, structural and deep learning integration

Predicting and quantifying phenotypic consequences of genetic variants in rare disorders is a major challenge, particularly pertinent for ‘actionable’ genes such as thyroid hormone transporter MCT8 (encoded by the X-linked SLC16A2 gene), where loss-of-function (LoF) variants cause a rare neurodevelopmental and (treatable) metabolic disorder in males. The combination of deep phenotyping data with functional and computational tests and with outcomes in population cohorts, enabled us to: (i) identify the genetic aetiology of divergent clinical phenotypes of MCT8 deficiency with genotype-phenotype relationships present across survival and 24 out of 32 disease features; (ii) demonstrate a mild phenocopy in ~400,000 individuals with common genetic variants in MCT8; (iii) assess therapeutic effectiveness, which did not differ among LoF-categories; (iv) advance structural insights in normal and mutated MCT8 by delineating seven critical functional domains; (v) create a pathogenicity-severity MCT8 variant classifier that accurately predicted pathogenicity (AUC:0.91) and severity (AUC:0.86) for 8151 variants. Our information-dense mapping provides a generalizable approach to advance multiple dimensions of rare genetic disorders.

Deep developmental phenotyping in children with tuberous sclerosis complex, with and without autism.

To characterize autism and co-occurring tuberous sclerosis-associated neuropsychiatric disorders (TAND) in children with tuberous sclerosis complex (TSC), addressing evidence gaps by using deep developmental phenotyping in a single cohort. This cohort study assessed autism characteristics, intelligence, adaptive function, language, and co-occurring conditions, using multidisciplinary direct assessment, in 50 children with TSC, comparing those with and without autism. Autistic children (28, 56%) had moderate mean scores for autistic characteristics (Autism Diagnostic Observation Schedule calibrated severity scores; social affect, 6.9 [standard deviation 2.5]; restricted and repetitive behaviour, 7.1 [standard deviation 2.3], but with considerable variation). Autistic children were more likely to be male (54% vs. 18%) and have intellectual disability (79% vs. 32%), language impairment (89% vs. 50%), executive dysfunction (70% vs. 29%), and externalizing behaviours (46% vs. 14%). Inattentive attention-deficit/hyperactivity disorder (ADHD) symptoms were frequent in autistic and not-autistic children (74% vs. 78%), and not influenced by intellectual ability. Language impairment occurred in 27% without autism or intellectual disability. TAND are complex and heterogenous in children with and without autism. Formal assessment of language function and ADHD symptoms should be considered in all children with TSC, regardless of autism categorization or intellectual ability. Language function needs greater consideration within TAND levels and clusters.

Schizotypy, Psychosis Proneness, and the Polygenic Risk for Schizophrenia and Resilience.

Schizotypy is a well-established phenotype for psychosis proneness and risk. Yet, its genetic underpinnings and relations to genetic bases of the schizophrenia spectrum are not well understood owing to conflicting findings. In a deep phenotyping approach, we hypothesized that genetic markers of risk for and to schizophrenia are differentially associated with (trait-level) dimensions of schizotypy and (state-level) prodromal symptoms. In 367 (130 male, 237 female) psychiatrically healthy young adults, we assessed multiple schizotypy instruments (OLIFE, SPQ-B, Multidimensional Schizotypy Scales), aggregated into composite scores, and a measure of prodromal symptoms (PQ-16). Those were tested for direct and interactive associations with the polygenic risk score (PRS) for schizophrenia and a novel PRS for resilience to schizophrenia. Both prodromal symptom number (rho = 0.16, pcorr = .018) and distress (rho = 0.14, pcorr = .027) were positively related to the schizophrenia PRS. Positive schizotypy showed a similar association but did not remain significant after correction (rho = 0.11, pcorr = .082). Schizophrenia PRS and disorganized schizotypy had a negative interactive effect on prodromal symptom distress (b = -0.10, pcorr = .048). The resilience score did not show any significant associations with any of the measures. These results further support the idea of a (partially) shared genetic basis of schizophrenia and nonclinical, predominantly positive expressions of the psychosis spectrum but also indicate relevant distinctions between the 2, possibly related to other modulating factors or general (transdiagnostic) psychopathological risk. In line with previous findings, effects seem to be more robust for state- than trait-level markers, but these may also be influencing each other.

Linking obesity-associated genotype to child language development: the role of early-life neurology-related proteomics and brain myelination.

Linking obesity-associated genotype to child language development: the role of early-life neurology-related proteomics and brain myelination.

Leveraging Transformers-based models and linked data for deep phenotyping in radiology.

Despite significant investments in the normalization and the standardization of Electronic Health Records (EHRs), free text is still the rule rather than the exception in clinical notes. The use of free text has implications in data reuse methods used for supporting clinical research since the query mechanisms used in cohort definition and patient matching are mainly based on structured data and clinical terminologies. This study aims to develop a method for the secondary use of clinical text by: (a) using Natural Language Processing (NLP) for tagging clinical notes with biomedical terminology; and (b) designing an ontology that maps and classifies all the identified tags to various terminologies and allows for running phenotyping queries. Transformers-based NLP Models, concretely pre-trained RoBERTa language models, were used to process radiology reports and annotate them identifying elements matching UMLS Concept Unique Identifiers (CUIs) definitions. CUIs were mapped into several biomedical ontologies useful for phenotyping (e.g., SNOMED-CT, HPO, ICD-10, FMA, LOINC, and ICPC2, among others) and represented as a lightweight ontology using OWL (Web Ontology Language) constructs. This process resulted in a Linked Knowledge Base (LKB), which allows running expressive queries to retrieve reports that comply with specific criteria using automatic reasoning. Although phenotyping tools mostly rely on relational databases, the combination of NLP and Linked Data technologies allows us to build scalable knowledge bases using standard ontologies from the Web of data. Our approach enables us to execute a pipeline which input is free text and automatically maps identified entities to a LKB that allows answering phenotyping queries. In this work, we have only used Spanish radiology reports, although it is extensible to other languages for which suitable corpora are available. This is particularly valuable in regional and national systems dealing with large research databases from different registries and cohorts and plays an essential role in the scalability of large data reuse infrastructures that require indexing and governing distributed data sources.

Deep plasma and tissue proteome profiling of knockout mice reveals pathways associated with Svep1 deficiency.

Deep plasma and tissue proteome profiling of knockout mice reveals pathways associated with Svep1 deficiency.

Systemic Identification of Functionally Conserved Long Noncoding RNA Metabolic Regulators in Human and Mouse Livers.

Systemic Identification of Functionally Conserved Long Noncoding RNA Metabolic Regulators in Human and Mouse Livers.

Deep phenotyping of patients with MASLD upon high-intensity interval training.

Exercise is a key component of lifestyle management in patients with metabolic dysfunction-associated steatotic liver disease (MASLD), but neither its therapeutic effect on the active stage of the disease, that is metabolic dysfunction-associated steatohepatitis (MASH) nor the mediating mechanisms have been characterized. Therefore, we performed multi-omic phenotyping of patients with MASLD-MASH on an exercise program. Fifteen patients with MASLD conducted high-intensity interval training (HIIT) combined with home-based training for 12 weeks. MASLD was evaluated using histology, transient elastography, and multiparametric magnetic resonance imaging (MRI) before and after the intervention. Change in maximal oxygen consumption (VO2max) and MRI-determined liver fat were compared with a control group of patients with MASLD (n= 22). RNA sequencing was performed on liver, muscle, and fat biopsies of patients in the exercise group. Stool was analyzed by shotgun metagenomics and untargeted metabolomics was performed on plasma, urine, adipose, and stool. HIIT increased VO2max by 10.1% and improved mitochondrial metabolism in skeletal muscle, indicating improved cardiorespiratory fitness and adherence. VO2max increased significantly in the exercise group compared with controls. Histologically, no reduction in steatosis, MASH, or liver fibrosis was observed; however, transient elastography tended to improve. MRI-determined liver fat did not change in the exercise group compared with controls. HIIT induced changes in mRNA expression of genes related to beiging of adipose tissue and fibrogenesis in liver. In addition, specific gut microbial taxa and metabolites changed. HIIT increased cardiorespiratory fitness and induced beneficial gene expression changes in muscle, adipose tissue, and liver, but without translation into histological improvement of MASLD. Longer exercise intervention trials are warranted to validate or refute current recommendations for exercise as a cornerstone treatment for MASLD-MASH. Despite exercise being considered as a key component of lifestyle management for steatotic liver disease, neither the clinical effects nor the mechanisms involved are completely understood. We show that a high-intensity interval training (HIIT) program in 15 patients with metabolic dysfunction-associated steatotic liver disease (MASLD) improved cardiorespiratory fitness, compared with 22 control patients with MASLD who did not participate in an exercise program, however, it did not improve MASLD. HIIT induced a positive effect on fat tissue and muscle metabolism which was accompanied with changes in certain gut bacteria and metabolites in blood and urine. These findings improve our understanding of the effects of exercise on the whole-body metabolism in relation to steatotic liver disease. As such, this study provides a basis for future exercise interventions in patients with MASLD, required to thoroughly test current guideline advice for exercise as a cornerstone treatment for MASLD of all stages. Dutch Trial Register (registration number NL7932).

Inherited Dyslipidemic Splenomegaly: A Genetic Macrophage Storage Disorder Caused by Disruptive Apolipoprotein E (APOE) Variants.

Persistent splenomegaly, often an incidental finding, can originate from a number of inherited metabolic disorders (IMDs). Variants of APOE are primarily known as risk factors in terms of cardiovascular disease; however, severe dysfunction of APOE can result in a disease phenotype with considerable overlap with lysosomal storage disorders (LSDs), including splenomegaly and gross elevation of N-palmitoyl-O-phosphocholine-serine (PPCS). A case study (deep phenotyping, genetic and FACS analysis) and literature study was conducted. The index patient, with a family history of early-onset cardiovascular disease, presented with splenic infarctions in a grossly enlarged spleen. The identified genetic cause was homozygosity for two APOE variants (c.604C>T, p.(Arg202Cys) and c.512G>A, p.(Gly171Asp); ε1/ε1), resulting in a macrophage storage phenotype resembling an LSD that was also present in the brother of the index patient. A FACS analysis of the circulating monocytes showed increased lipid content and the expression of activation markers (CD11b, CCR2, CD36). This activated state enhances lipoprotein intake, which eventually converts these monocytes/macrophages into foam cells, accumulating in tissues (e.g., spleen and vascular wall). A literature search identified seven individuals with splenomegaly caused by APOE variants (deletion of leucine at position 167). The combined data from all patients identified male gender, splenectomy and obesity as potential modifiers determining the severity of the phenotype (i.e., degree of triglyceride increase in plasma and/or spleen size). Symptoms are (partially) reversible by lipid-lowering medication and energy restricted diets and splenectomy is contra-indicated. Inherited dyslipidemic splenomegaly caused by disruptive APOE variants should be included in the differential diagnoses of unexplained splenomegaly with abnormal lipid profiles. A plasma lipid profile consistent with dysbetalipoproteinemia is a diagnostic biomarker for this IMD.

Expanding the mutational spectrum of ReNU syndrome: insights into 5' Stem-loop variants.

A narrow spectrum of heterozygous variants in RNU4-2, encoding the small nuclear RNA (snRNA) U4, underlies ReNU syndrome, a neurodevelopmental disorder (NDD) characterized by moderate to severe developmental delay (DD), intellectual disability (ID), a distinctive facial gestalt, and multisystem involvement. Pathogenic variants have primarily been reported within an 18-nt critical region contributing to stabilizing the U4/U6 snRNA duplex and proper spliceosome assembly. By combining whole genome sequencing reanalysis and targeted direct sequencing in 190 molecularly unexplained NDD cases, we report on five affected individuals carrying pathogenic/putative pathogenic RNU4-2 variants (2.6%). Three individuals harbored the recurrent pathogenic n.64_65insT variant, while two were heterozygous for private/rare maternally inherited variants (n.30 A > T and n.43_44insT) within the 5' Stem-loop region. Deep clinical phenotyping confirmed a homogeneous constellation of features in all individuals, with global DD, ID, brain malformations, and a recognizable facial gestalt representing core findings. Based on structural homology models and available cryo-EM data, n.30 A > T and n.43_44insT were predicted to disrupt key intra- and inter-molecular interactions critical for spliceosome function. Our findings expand the mutational spectrum of ReNU syndrome, and confirm the 5' Stem-loop as a second mutational hotspot in RNU4-2. We propose that a more complex genetics likely underlies the inheritance of a subset of disease-causing RNU4-2 variants from an apparently unaffected parent. We anticipate a relatively high proportion of pathogenic RNU4-2 variants among individuals with unclassified NDD despite extensive genomic testing, and propose a set of facial gestalt core features as a clinical screening tool to prioritize patients for RNU4-2 analysis.