Biology subjects

Pontikos, N.

Publications and source records attributed to Pontikos, N..

13 recordsLinked to original sources

bioRxiv2025

GIMMEcpg: Global Imputation of Mean CpG MEthylation in Real-time

Whole-genome DNA methylation (methylome) analysis is of broad interest to biomedical research due to its central role in human development and disease. However, generating high-quality methylomes at scale remains challenging due to inherent technical limitations. While imputation has the potential to help overcome this problem, no existing approach adequately addresses the scaling issue. Here, we present GIMMEcpg (Global Imputation of Mean cpg MEthylation), a novel imputation tool that scales efficiently from single samples to large cohort studies. GIMMEcpg uses a custom feature dataset built from known CpG sites within the same dataset to impute missing values by calculating the distance-weighted mean of the methylation value of the two immediately neighbouring CpG sites. We benchmarked GIMMEcpg for speed and accuracy against multiple imputation methods using downsampled datasets produced from high-quality ([~]100x) Whole Genome Bisulfite Sequencing (WGBS) data. With a 10x downsampled dataset, GIMMEcpg was able to process the dataset and impute 9.14 Million CpG sites within 7 seconds (R: 0.78, MAE: +5.6%, RMSE: +10.9%). Our results demonstrate that GIMMEcpg is 39-2,562 times faster than three existing methylation imputation tools (BoostMe, DeepCpG, and MethImpute) while maintaining comparable accuracy. To quantify GIMMEcpgs scalability, we applied it to the most extensive single collection of WGBS data (N=645 at variable coverage) from the EpiATLAS generated by the International Human Epigenome Consortium (IHEC). Using a single, standard CPU server, GIMMEcpg processed and imputed an additional 2.4 billion CpG methylation values across the 645 datasets in less than a day, enriching the EpiATLAS methylome resource by 20%. This demonstrates that GIMMEcpg scales to large cohort studies with only a subtle impact on accuracy, as illustrated by our benchmark. We also developed a machine learning variant, GIMMEcpg.ml, which delivers a higher accuracy compared to existing methodologies. Using the same 10x downsampled benchmarking dataset, GIMMEcpg.ml achieved a Person Correlation of 0.87 compared to the ground truth, representing an improvement of 0.11 over the best performing alternative method. Additionally, GIMMEcpg.ml has a Mean Absolute Error (MAE) of 8.67%, which is 2.63% lower than the most accurate performing alternative. While this enhanced accuracy comes at the cost of increased computation requirements, GIMMEcpg.ml is a useful tool where higher accuracy is preferred over scalability. For increased accessibility, GIMMEcpg is freely available under an MIT license as R and Python packages at https://github.com/ucl-medical-genomics/gimmecpg-r and https://github.com/ucl-medical-genomics/gimmecpg-python, respectively.

bioinformatics↗

bioRxiv2024

Inhibitory KIRs decrease HLA class II-mediated protection in Type 1 Diabetes

Inhibitory killer cell immunoglobulin-like receptors (iKIRs) are a family of inhibitory receptors that are expressed by natural killer cells and late-stage differentiated T cells. There is accumulating evidence that iKIRs regulate T cell-mediated immunity. Recently, we reported that T cell-mediated control was enhanced by iKIRs in chronic viral infections. We hypothesized that in the context of autoimmunity, where an enhanced T cell response might be considered detrimental, iKIRs would have an opposite effect. We studied Type 1 diabetes (T1D) as a paradigmatic example of autoimmunity. In T1D, variation in the Human Leucocyte Antigen (HLA) genes explains up to 50% of the genetic risk, indicating that T cells have a major role in T1D etiopathogenesis. To investigate if iKIRs affect this T cell response we asked whether HLA associations were modified by iKIR genes. We conducted an immunogenetic analysis of a case-control T1D dataset (N= 11,961) and found that iKIR genes, in the presence of genes encoding their ligands, have a consistent and significant effect on protective HLA class II genetic associations. Our results were validated in an independent data set. We conclude that iKIRs significantly decrease HLA class II protective associations and suggest that iKIRs regulate CD4+ T cell responses in T1D.

immunology↗

bioRxiv2023

Deciphering novel TCF4-driven molecular origins and mechanisms underlying a common triplet repeat expansion-mediated disease

Fuchs endothelial corneal dystrophy (FECD) is an age-related cause of vision loss, and the most common repeat expansion-mediated disease in humans characterised to date. Up to 80% of European FECD cases have been attributed to expansion of a non-coding CTG repeat element (termed CTG18.1) located within the ubiquitously expressed transcription factor encoding gene, TCF4. The non-coding nature of the repeat and the transcriptomic complexity of TCF4 have made it extremely challenging to experimentally decipher the molecular mechanisms underlying this disease. Here we comprehensively describe CTG18.1 expansion-driven molecular components of disease within primary patient-derived corneal endothelial cells (CECs), generated from a large cohort of individuals with CTG18.1-expanded (Exp+) and CTG 18.1-independent (Exp-) FECD. We employ long-read, short-read, and spatial transcriptomic techniques to interrogate expansion-specific transcriptomic biomarkers. Interrogation of long-read sequencing and alternative splicing analysis of short-read transcriptomic data together reveals the global extent of altered splicing occurring within Exp+ FECD, and unique transcripts associated with CTG18.1-expansions. Similarly, differential gene expression analysis highlights the total transcriptomic consequences of Exp+ FECD within CECs. Furthermore, differential exon usage, pathway enrichment and spatial transcriptomics reveal TCF4 isoform ratio skewing solely in Exp+ FECD with potential downstream functional consequences. Lastly, exome data from 134 Exp- FECD cases identified rare (minor allele frequency <0.005) and potentially deleterious (CADD>15) TCF4 variants in 7/134 FECD Exp- cases, suggesting that TCF4 variants independent of CTG18.1 may increase FECD risk. In summary, our study supports the hypothesis that at least two distinct pathogenic mechanisms, RNA toxicity and TCF4 isoform-specific dysregulation, both underpin the pathophysiology of FECD. We anticipate these data will inform and guide the development of translational interventions for this common triplet-repeat mediated disease. Authors summaryFuchs endothelial corneal dystrophy (FECD) leads to vision loss and is the most common repeat expansion-mediated disease characterised to date; most individuals with FECD harbour a non-coding CTG repeat expansion within the gene TCF4. FECD represents an important paradigm for other rare and devastating neurological repeat-mediated diseases, given its relatively mild and tissue-specific nature. Intriguingly, despite TCF4 being ubiquitously expressed, individuals with FECD only experience corneal disease, and the biological reason for this tissue-specificity remains elusive. Here, we use tissue from 31 individuals with FECD to perform complementary long-read, short-read and spatial transcriptomic analyses to enhance our understanding of mechanisms underpinning this disease. These data highlight that at least two mechanisms, RNA toxicity and TCF4 isoform dysregulation, underlie the disease state in affected corneal cells. Furthermore, TCF4 isoform skewing, with evidence of downregulation, suggests this mechanism in part may explain the unique vulnerability of the cornea. In addition, 7/134 FECD expansion negative cases were identified to harbour rare and potentially deleterious TCF4 variants, further supporting the hypothesis that dysregulation of TCF4 may be key to FECD pathophysiology. Biological insights presented here will guide the development of personalised FECD therapies and may inform the development of repeat-expansion mediated therapies more broadly.

molecular biology↗

bioRxiv2019

Frequency and Distribution of Corneal Astigmatism and Keratometry Features: Methodology and Findings of the UK Biobank Study

PurposeTo describe corneal astigmatism in the UK Biobank population, to look for associations with other biometric variables and socio-demographic factors, and to report the proportion with abnormal keratometry and irregular astigmatism suggestive of pathological corneal ectasias such as keratoconus.\n\nMethodsCross-sectional data were obtained from UK Biobank (www.ukbiobank.ac.uk/). A subsample of 107,452 participants from UK communities had undergone an enhanced ophthalmic examination including autorefractor keratometry (Tomey RC 5000, Tomey Corp., Nagoya, Japan). After quality control and applying relevant exclusions, data on corneal astigmatism on 83,751 participants was available for analysis. Potential associations were tested through univariable regression and significant parameters carried forward for multivariable analysis.\n\nResultsIn a univariable analysis, the characteristics significantly protective against corneal astigmatism were gender (male), older age, darker skin colour and increased alcohol intake (all p<0.001). The parameters significantly associated with increased corneal astigmatism were older age at completion of full time education, use of UV protection and lower corrected visual acuity. After inclusion in the multivariable analysis, age, gender, age at completion of full time education, corrected visual acuity and skin colour remained significant (all p<0.001). Increased corneal astigmatism was also found to be significantly associated with amblyopia or strabismus. No individuals with abnormal keratometry or irregular astigmatism were reported.\n\nConclusionsThis analysis of associations with astigmatism in a large cohort of volunteers confirms previous associations including adverse associations with younger age and female gender, and identified novel associations including darker skin colour and frequency of alcohol intake. The highest risk group for corneal astigmatism were younger females of lighter skin colour, having completed full time education later, with higher logMAR corrected visual acuity. We also confirmed that corneal astigmatism is a high risk factor for amblyopia and strabismus. Finally since no cases of keratoconus were identified, this would suggest that simple keratometry indices may not be sufficient for population screening of keratoconus.

epidemiology↗

bioRxiv2019

Feasibility of Automated Deep Learning Design for Medical Image Classification by Healthcare Professionals with Limited Coding Experience

Deep learning has huge potential to transform healthcare. However, significant expertise is required to train such models and this is a significant blocker for their translation into clinical practice. In this study, we therefore sought to evaluate the use of automated deep learning software to develop medical image diagnostic classifiers by healthcare professionals with limited coding - and no deep learning - expertise.\n\nWe used five publicly available open-source datasets: (i) retinal fundus images (MESSIDOR); (ii) optical coherence tomography (OCT) images (Guangzhou Medical University/Shiley Eye Institute, Version 3); (iii) images of skin lesions (Human against Machine (HAM)10000) and (iv) both paediatric and adult chest X-ray (CXR) images (Guangzhou Medical University/Shiley Eye Institute, Version 3 and the National Institute of Health (NIH)14 dataset respectively) to separately feed into a neural architecture search framework that automatically developed a deep learning architecture to classify common diseases. Sensitivity (recall), specificity and positive predictive value (precision) were used to evaluate the diagnostic properties of the models. The discriminative performance was assessed using the area under the precision recall curve (AUPRC). In the case of the deep learning model developed on a subset of the HAM10000 dataset, we performed external validation using the Edinburgh Dermofit Library dataset.\n\nDiagnostic properties and discriminative performance from internal validations were high in the binary classification tasks (range: sensitivity of 73.3-97.0%, specificity of 67-100% and AUPRC of 0.87-1). In the multiple classification tasks, the diagnostic properties ranged from 38-100% for sensitivity and 67-100% for specificity. The discriminative performance in terms of AUPRC ranged from 0.57 to 1 in the five automated deep learning models. In an external validation using the Edinburgh Dermofit Library dataset, the automated deep learning model showed an AUPRC of 0.47, with a sensitivity of 49% and a positive predictive value of 52%. The quality of the open-access datasets used in this study (including the lack of information about patient flow and demographics) and the absence of measurement for precision, such as confidence intervals, constituted the major limitation of this study.\n\nAll models, except for the automated deep learning model trained on the multi-label classification task of the NIH CXR14 dataset, showed comparable discriminative performance and diagnostic properties to state-of-the-art performing deep learning algorithms. The performance in the external validation study was low. The availability of automated deep learning may become a cornerstone for the democratization of sophisticated algorithmic modelling in healthcare as it allows the derivation of classification models without requiring a deep understanding of the mathematical, statistical and programming principles. Future studies should compare several application programming interfaces on thoroughly curated datasets.

bioinformatics↗

bioRxiv2019

The Moorfields AMD Database Report 2 - Fellow Eye Involvement with Neovascular Age-related Macular Degeneration

Background/AimsNeovascular age-related macular degeneration (nAMD) is frequently bilateral, and previous reports on fellow eyes have assumed sequential treatment after a period of treatment of the first eye only. The aim of our study was to analyse baseline characteristics and visual acuity (VA) outcomes of fellow eye involvement with nAMD, specifically differentiating between sequential and non-sequential (due to macular scarring in the first eye) anti-vascular endothelial growth factor treatment and timelines for fellow eye involvement.\n\nMethodsRetrospective, electronic medical record database study of the Moorfields AMD database of 8174 eyes/120,756 single entries with data extracted between October 21, 2008 and August 9, 2018. The dataset for analysis consisted of 1180 sequential, 413 nonsequential, and 1110 unilateral eyes.\n\nResultsMean VA of sequentially treated fellow eyes at baseline was significantly higher (62{+/-}13), VA gain over two years lower (0.65{+/-}14), and proportion of eyes with good VA ([≥]20/40 or 70 letters) higher (46%) than the respective first eyes (baseline VA 54{+/-}16, VA gain at two years 5.6{+/-}15, percentage of eyes with good VA 38%). Non-sequential fellow eyes showed baseline characteristics and VA outcomes similar to first eyes. Fellow eye involvement rate was 32% at two years, and median time interval to fellow eye involvement was 71 (IQR 27-147) weeks.\n\nConclusionThis reports shows sequentially treated nAMD fellow eyes have better baseline and final VA than non-sequentially treated eyes after 2 years of treatment. Sequentially treated eyes also had a greater proportion with good VA after 2 years of treatment.\n\nPRECISDepending on age, fellow eye involvement occurs in 32% of patients with neovascular AMD by two years. Fellow eyes generally maintain better vision, except in cases where late-stage disease in the first eye was untreated.

epidemiology↗

bioRxiv2018

One and Two Year Visual Outcomes from the Moorfields AMD Database - an Open Science Resource for the Study of Neovascular Age-related Macular Degeneration

ObjectivesTo analyse treatment outcomes and share clinical data from a large, single-center, well-curated database (8174 eyes / 6664 patients with 120,756 single entries) of patients with neovascular age related macular degeneration (AMD) treated with anti-vascular endothelial growth factor (VEGF). By making our depersonalised raw data openly available, we aim to stimulate further research in AMD, as well as setting a precedent for future work in this area.\n\nSettingRetrospective, comparative, non-randomised electronic medical record (EMR) database cohort study of the UK Moorfields AMD database with data extracted between 2008 and 2018.\n\nParticipants3357 eyes/patients (61% female). Extraction criteria were [≥] 1 ranibizumab or aflibercept injection, entry of \"AMD\" in the diagnosis field of the EMR, and a minimum of one year of follow-up. Exclusion criteria were unknown date of first injection and treatment outside of routine clinical care at Moorfields before the first recorded injection in the database.\n\nMain outcome measuresPrimary outcome measure was change in VA at one and two years from baseline as measured in Early Treatment Diabetic Retinopathy Study (ETDRS) letters. Secondary outcomes were the number of injections and predictive factors for VA gain.\n\nResultsMean VA gain at one-year and two years were +5.5{+/-}0.5 and +4.9{+/-}0.68 letters respectively. Fifty-four percent of eyes gained [≥]5 letters at two years, 63% had stable VA ({+/-}[≤]14 letters), forty-four percent of eyes maintained good VA ([≥]70 letters). Patients received a mean of 7.7{+/-}0.06 injections during year one and 13.0{+/-}0.2 injections over two years.\n\nYounger age, lower baseline VA, and more injections were associated with higher VA gain at two years.\n\nConclusionThis study benchmarks high quality EMR study results of real life AMD treatment and promotes open science in clinical AMD research by making the underlying data publicly available.\n\nStrengths and limitations of this study- Large sample size, retrospective, single centre, electronic medical record database study\n- High quality real life data\n- Open science approach with sharing of depersonalised raw data

epidemiology↗

bioRxiv2018

Phenogenon: Gene to Phenotype Associations for Rare Genetic Diseases

As genome sequencing is increasingly applied to molecular diagnosis of rare Mendelian disorders, large number of patients with diverse phenotypes have their genomic and phenotypic data pooled together to uncover new genotype - phenotype relations. We introduce Phenogenon, a method that combines: the power of Human Phenotype Ontology for describing patient phenotypes, gnomAD for estimating rare variant population frequency, and CADD for variant pathogenicity prediction. By using a divide and conquer approach, we demonstrate here that Phenogenon is able to uncover true gene to phenotype relations, such as \"ABCA4 - Macular dystrophy\" and \"SCN1A - Seizures\". Additionally, it accurately infers mode of inheritance, such as a recessive mode of inheritance in the case of the \"ABCA4 - Macular dystrophy\" relationship and a dominant mode of inheritance with the \"SCN1A - Seizures\" relationship. We also found that CADD has more power to detect early-onset rare genetic diseases than late-onset diseases. In this study, we ran Phenogenon against a diverse cohort of 3288 patients. Among the top 13 gene-phenotype relations, seven were previously known. We also highlight four potentially novel gene - phenotype relations such as \"SIPA1L3 - Abnormal electroretinogram\".

bioinformatics↗

bioRxiv2018

PGP-UK: a research and citizen science hybrid project in support of personalized medicine

Molecular analyses such as whole-genome sequencing have become routine and are expected to be transformational for future healthcare and lifestyle decisions. Population-wide implementation of such analyses is, however, not without challenges, and multiple studies are ongoing to identify what these are and explore how they can be addressed. Defined as a research project, the Personal Genome Project UK (PGP-UK) is part of the global PGP network and focuses on open data sharing and citizen science to advance and accelerate personalized genomics and medicine. Here we report our findings on using an open consent recruitment protocol, active participant involvement, open access release of personal genome, methylome and transcriptome data and associated analyses, including 47 new variants predicted to affect gene function and innovative reports based on the analysis of genetic and epigenetic variants. For this pilot study, we recruited ten participants willing to actively engage as citizen scientists with the project. In addition, we introduce Genome Donation as a novel mechanism for openly sharing previously restricted data and discuss the first three donations received. Lastly, we present GenoME, a free, open-source educational app suitable for the lay public to allow exploration of personal genomes. Our findings demonstrate that citizen science-based approaches like PGP-UK have an important role to play in the public awareness, acceptance and implementation of genomics and personalized medicine.

genomics↗

bioRxiv2017

Pheno4J: A Gene To Phenotype Graph Database

SummaryEfficient storage and querying of large amounts of genetic and phenotypic data is crucial to contemporary clinical genetic research. This introduces computational challenges for classical relational databases, due to the sparsity and sheer volume of the data. Our Java based solution loads annotated genetic variants and well phenotyped patients into a graph database to allow fast efficient storage and querying of large volumes of structured genetic and phenotypic data. This abstracts technical problems away and lets researchers focus on the science rather than the implementation. We have also developed an accompanying webserver with end-points to facilitate querying of the database.\n\nAvailability and ImplementationThe Java code and python code is available at https://github.com/phenopolis/pheno4i\n\nContactn.pontikos@ucl.ac.uk

bioinformatics↗

bioRxiv2017

The human oral microbiome is shaped by shared environment rather than genetics: evidence from a large family of closely-related individuals

The human microbiome is affected by multiple factors, including the environment and host genetics. In this study, we analyzed the oral microbiome of an extended family of Ashkenazi Jewish individuals living in several cities and investigated associations with both shared household and host genetic similarities. We found that environmental effects dominated over genetic ones. While there was weak evidence of geographic structuring at the level of cities, we observed a large and significant effect of shared household on microbiome composition, supporting the role of immediate shared environment in dictating the presence or absence of taxa. This effect was also seen when including adults who had grown up in the same household but moved out prior to the time of sampling, suggesting that the establishment of the oral microbiome earlier in life may affect its long-term composition. We found weak associations between host genetic relatedness and microbiome dissimilarity when using family pedigrees as proxies for genetic similarity. However this association disappeared when using more accurate measures of kinship based on genome-wide genetic markers, indicating that environment rather than host genetics is the dominant factor affecting the composition of the oral microbiome in closely-related individuals. Our results support the concept that there is a consistent core microbiome conserved across global scales, but that small-scale effects due to shared living environment significantly affect microbial community composition.\n\nIMPORTANCEPrevious research shows that relatives have a more similar oral microbiome composition than non-relatives, but it remains difficult to distinguish the effects of relatedness and shared household environment. Furthermore, pedigree measures may not accurately measure host genetic similarity. In this study, we include genetic relatedness based on genome-wide SNPs (rather than pedigree measures) and shared environment in the same analysis. We quantify the relative importance of these factors by studying the oral microbiome in members of a large extended Ashkenazi Jewish family who share a similar diet and lifestyle despite living in different locations. We find that host genetics plays no significant role and that the dominant factor is shared environment at the household level. We also find that this effect appears to persist in individuals who have moved out of the parental household, suggesting that the oral microbiome established earlier in life persists long-term.

microbiology↗

bioRxiv2017

Duplication events downstream of IRX1 cause North Carolina macular dystrophy at the MCDR3 locus

Autosomal dominant North Carolina macular dystrophy (NCMD) is believed to represent a failure of macular development. The disorder has been assigned by linkage to two loci, MCDR1 on chromosome 6q16 and MCDR3 on chromosome 5p15-p13. Recently, noncoding variants upstream of PRDM13 and a large duplication including IRX1 have been identified. However, the underlying disease-causing mechanism remains uncertain. Through a combination of sequencing studies, we report two novel overlapping duplications at the MCDR3 locus, in a gene desert downstream of IRX1 and upstream of ADAMTS16. One duplication of 43 kb was identified in nine NCMD families (with evidence for a shared ancestral haplotype), and another one of 45 kb was found in a single family. The MCDR3 locus is thus refined to a shared region of 39 kb that contains DNAse hypersensitive sites active at a restricted time window during retinal development. Publicly available data confirmed expression of IRX1 and ADAMTS16 in human fetal retina, with IRX1 preferentially expressed in fetal macula. These findings represent a major advance in our understanding of the molecular genetics of NCMD at the MCDR3 locus and provide insights into the genetic pathways involved in human macular development.\n\nAbbreviations list

genomics↗

bioRxiv2016

Phenopolis: an open platform for harmonization and analysis of sequencing and phenotype data

SummaryPhenopolis is an open-source web server which provides an intuitive interface to genetic and phenotypic databases. It integrates analysis tools which include variant filtering and gene prioritisation based on phenotype. The Phenopolis platform will accelerate clinical diagnosis, gene discovery and encourage wider adoption of the Human Phenotype Ontology in the study of rare disease.\n\nAvailability and ImplementationA demo of the website is available at http://phenopolis.github.io (username: demo, password: demo123). If you wish to install a local copy, souce code and installation instruction are available at https://github.com/pontikos/phenopolis. The software is implemented using Python, MongoDB, HTML/Javascript and various bash shell scripts.\n\nContactn.pontikos@ucl.ac.uk\n\nSupplementary informationhttp://phenopolis.github.io

bioinformatics↗