Open Data Institute

Building and sustaining High-Performance Computing (HPC) infrastructure for bioinformatics research in resource-limited settings presents significant technical, financial and operational challenges. Institutions in low-and middle-income regions often face constraints such as limited technical expertise, unstable infrastructure and restricted funding which can hinder the deployment of large-scale computational platforms necessary for modern genomics and bioinformatics analyses. We present a scalable and modular HPC framework developed at the Uganda Virus Research Institute (UVRI) to support large-scale genomics and other omics data analyses in resource-limited settings. The framework integrates open-source HPC management tools, infrastructure automation, and reproducible configuration management to enable reliable deployment and maintenance. Optimized storage and networking configurations combined with a phased capacity-building strategy support high-throughput genomic workflows while strengthening local technical expertise. From our implementation experience, we derive ten practical design and operational rules that provide a transferable methodology for establishing and sustaining in-house HPC infrastructure. These rules emphasize strategic investment in human capacity, structured planning, leveraging collaborations, adoption of open-source technologies and service management practices to improve operational resilience and long-term sustainability. The design principles, automation strategies and implementation guidelines described in this work are applicable to institutions seeking to establish sustainable HPC resources for bioinformatics research in resource-constrained environments.

This article reports the results of the second iteration of the autoPET challenge on automated lesion segmentation in whole-body PET/CT, held in conjunction with the 26th International Conference on Medical Image Computing and Computer Assisted Intervention in 2023. In contrast to the first autoPET challenge, which served as a proof of concept, this study investigates whether machine learning-based segmentation models trained on data from a single source can maintain performance across clinically relevant variations in PET/CT data, reflecting the demands of real-world deployment. Methods: A comprehensive biomedical segmentation challenge on PET/CT domain generalization was designed and conducted. Participants were tasked to train machine learning models on annotated whole-body 18F-FDG data (n = 1,014). These models were then evaluated on a test set of 200 samples from 5 clinically relevant domains, including variations in institutions, pathologies, and populations and a different tracer. Performance was measured in terms of average dice similarity coefficient, average false-positive volume, and average false-negative volume. The best-performing teams were awarded in 3 categories. Furthermore, a detailed analysis was conducted after the challenge, examining results across domains and unique instances, along with a ranking analysis. Results: Generalization from a single-source domain remains a significant challenge. Seventeen international teams successfully participated in the challenge. The best-performing team reached an average dice similarity coefficient of 0.5038, a mean false-positive volume of 87.8388 mL, and a mean false-negative volume of 8.4154 mL on the test set. nnU-Net was the most commonly used framework, with most participants using a 3-dimensional U-Net. Despite competitive in-domain results, out-of-domain performance deteriorated substantially, particularly on pediatric and prostate-specific membrane antigen data. Detailed error analysis revealed frequent false-positives due to physiologic uptake and decreased sensitivity in detecting small or low-uptake lesions. A majority-vote ensemble offered minimal performance gains, whereas an oracle ensemble indicates hypothetical gains. Ranking analysis showed no single team consistently outperformed all others across ranking schemes. Conclusion: The second autoPET challenge provides a comprehensive evaluation of the current state of automated PET/CT tumor segmentation, highlighting both progress and persistent challenges of single-source domain generalization and the need for diverse public datasets to enhance algorithm robustness.

<h3></h3> Heterozygous variants in RNU4-2, a small nuclear RNA (snRNA) component of the major spliceosome, have recently been shown to cause ReNU syndrome, a highly prevalent neurodevelopmental disorder (NDD). Recessive spliceosomopathies have only been described in the minor spliceosome snRNAs RNU12 and RNU4ATAC. Here, we describe a novel recessive NDD caused by biallelic variants in RNU4-2. Through international collaborations, we identify a cohort of 35 individuals and obtain detailed clinical information for 26 individuals. The disorder is characterised by global developmental delay (26/26, 100%), pronounced speech delay (absent speech in 15/26, 58%), hypotonia (20/24, 83%), and seizures (16/26, 62%). We observe a striking phenotype of progressive microcystic lesions in the periventricular white matter in 10/13 individuals (77%) with brain MRIs. Pathogenic variants cluster outside of the ReNU syndrome ‘critical regions’, within the Stem II region, k-turn region, and Sm protein binding site of RNU4-2. These variants are absent from UK Biobank in the homozygous (N=490,541) or compound heterozygous (N=200,011) states. In a saturation genome editing assay of RNU4-2, variants are significantly depleted for 5/7 (71%) homozygous genotypes and 9/13 (69%) compound heterozygous genotypes (at least one allele). In summary, this novel disorder expands the clinical and genetic spectrum of RNU4-2-related disease.

Abstract While much progress has been made in reducing the incidence of HIV-1 infection in sub-Saharan Africa in recent years, bringing the epidemic to an end will require identification of the demographic groups that continue to contribute to transmission. Pathogen phylogenetics and individual-based mathematical models (IBMs) of transmission are approaches that enable researchers to explore such questions. Here, we used both methods to characterise the ages and sexes of the individuals involved in heterosexual transmission in the context of the HPTN 071 (PopART) trial in Zambia. The results were concordant, and show that the male partner was on average older than the female by less than seven years, with larger age gaps in male-to-female than female-to-male transmissions. We found that the largest gaps for female recipients were amongst the youngest of those recipients. Conversely, the youngest male recipients saw the smallest gaps. We further used the IBM to demonstrate that transmission to new age cohorts first entering into sexual activity is driven predominantly by male-to-female transmission. We also simulated the PopART universal testing and treatment intervention into the future to show that effective treatment of under-35-year-olds accounts for 93.8% of the reduction in incidence by 2039, while effective treatment of under-35-year-old men accounts for 62.1%. Finally, we simulated a one-year cessation of ART treatment for the whole population, which resulted in an immediate increase in the average age at transmission of both sources and recipients. With it becoming ever more expensive and difficult to find treatment-naive individuals and link them to care, targeted interventions for demographic groups such as under-35 men may be the key to finally ending HIV.

ObjectivesThe ‘tumour, node, metastasis’ (TNM) classification of colorectal cancer (CRC) predicts prognosis and so is vital to consider in analyses of patterns and outcomes of care when using electronic health records. Unfortunately, it is often only available in free-text reports. This study aimed to develop regex-based text-processing algorithms that identify the reports describing CRC and extract the TNM staging at a low computational cost.MethodsThe CRC and TNM extraction algorithms were iteratively developed using 58 634 imaging and pathology reports of patients with CRC from the Oxford University Hospitals (OUH) and Royal Marsden (RMH) NHS Foundation Trusts (FT), with additional input from Imperial College Healthcare and Christie NHS FTs. The algorithms were evaluated on a stratified random sample of 400 OUH development data reports and 400 newer ‘unseen’ OUH reports. The reports were annotated with the help of two clinicians.ResultsThe CRC algorithm achieved at least 93.0% positive predictive value (PPV), 72.1% sensitivity, 64.0% negative predictive value (NPV) and 90.1% specificity for primary CRC on pathology reports. On imaging reports, it demonstrated at least 78.0% PPV, 91.8% sensitivity, 93.0% NPV and 80.9% specificity. For the main T/N/M categories, the TNM algorithm achieved PPVs of at least 93.9% (T), 97.7% (N) and 97.2% (M), and sensitivities of 63.6% (T), 89.6% (N) and 64.8% (M). NPVs were at least 45.0% (T), 91.1% (N), 88.4% (M), and specificities 95.7% (T), 98.1% (N), 99.3% (M). Reductions in performance were mostly due to implicit staging. For extracting explicit TNM stages, current or historical, the algorithm made no errors on 400 pathology reports and six errors on 400 imaging reports.ConclusionThe TNM algorithm accurately extracts explicit TNM staging, but other methods are needed for retrieving implicit stages. The CRC algorithm is accurate on non-supplementary reports, but outputs need additional review if higher precision is required.

Asian hepatointestinal schistosomiasis due to Schistosomajaponicum is prevalent in the Philippines and in Indonesia, while it is close to elimination in China. The second Asian schistosome, S.mekongi, is found in Cambodia and Laos. The main pathology caused by both species is liver fibrosis, which can cause significant morbidity and mortality, mainly due to portal hypertension leading to bleeding from esophageal varices. Ultrasonography was introduced several decades ago as a safe, fast, non-invasive, and relatively inexpensive technique for assessing chronic schistosomiasis-related hepatic pathology in the clinical and field settings. A standardized ultrasound protocol had been established by experts at a WHO-chaired meeting in Cairo, Egypt, in 1990. The peculiarities of sonomorphologic abnormalities caused by S.japonicum and S.mekongi were not sufficiently covered in the Cairo protocol and not addressed at all in the subsequent WHO chaired meeting in Niamey 1996. At a follow-up WHO-chaired meeting in Phnom Pehnh, Cambodia, in 2002, an attempt was made to develop a protocol for Asian schistosomiasis, but a protocol resulting from this meeting has never been published. Although several studies investigated the use of ultrasonography to assess S.japonicum- and S.mekongi-related sonomorphological morbidity across endemic areas the lack of a standardized protocol hampered the characterization of sonomorphologic abnormalities with regard to progression, reversibility, prognosis, and correlation to morbidity. In addition, the comparison of data from different endemic areas and populations remained difficult. Therefore, a WHO-chiared expert meeting took place in Basel, Switzerland in September 2024 with the aim to establish a standardized ultrasound protocol for reporting the pathology caused by S.japonicum and S.mekongi. The proposed protocol is described in this article.Graphical Supplementary InformationThe online version contains supplementary material available at 10.1186/s40249-025-01349-x.

Abstract The SARS-CoV-2 pandemic had a large impact on social mixing in the UK. This study analysed data from the Office for National Statistics Coronavirus Infection Survey to examine changes in contact patterns and self-reported symptoms through three winter seasons from 2020 to 2023. Using Generalised Additive Models, we estimated levels of various contacts over time, accounting for age, sex, ethnicity, and deprivation percentile, and compared these to trends in self-reported symptoms. Our estimates indicated steady increases in physical contacts from quarter-4 2020 to the end of the study in quarter-1 2023, with notable variation in age-specific trends. School closures and holiday periods had substantial impacts on contact patterns, particularly for children. Prevalence of reported symptoms also increased steadily over time, but varied much more within-season than most contacts; specifically, the relative increase in respiratory symptom prevalence during winter peaks between seasons was much larger than increases in contacts. Our estimates also suggested that while age played a crucial role in both contact patterns and symptom reporting, the effects of deprivation were less clear and far smaller. Our findings provide insights into changes in behaviours and symptoms during the pandemic, which may help inform future public health policy and infection modelling.

The use of whole-genome sequencing to monitor bacterial pathogens has provided crucial insights into their within-host evolution, revealing mutagenic and selective processes driving the emergence of antibiotic resistance, immune evasion phenotypes and adaptations that enable sustained human-to-human transmission. Deep genomic and metagenomic sequencing of intra-host pathogen populations is also enhancing our ability to track bacterial transmission, a key component of infection control. This Review discusses the major processes driving bacterial evolution within humans, including both pathogenic and commensal species. Initially, mutational processes, including how mutational signatures reveal pathogen biology, and the selective pressures driving evolution are considered. The dynamics of horizontal gene transfer and intra-host pathogen competition are also examined, followed by a focus on the emergence of bacterial pathogenesis. Finally, the Review focuses on the importance of within-host genetic diversity in tracking bacterial transmission and its implications for infectious disease control and public health.