Biology subjects

Goh, V.

Publications and source records attributed to Goh, V..

5 recordsLinked to original sources

bioRxiv2025

MCUR1-CCDC90B complex is a conserved regulator of metabolic homeostasis

The mitochondrial calcium uniporter regulator 1 (MCUR1) is an evolutionarily conserved protein of the inner mitochondrial membrane1, yet its physiological role has remained elusive. Although initially proposed to function as a subunit of the mitochondrial calcium uniporter complex (MCUC)2-4, emerging evidence suggests that MCUR1 has a broader functional spectrum5-7. Here, we identify a biallelic loss-of-function MCUR1 variant (c.802C>T; p.R268X) in a patient with a progressive neurological phenotype. This mutation leads to loss of MCUR1 protein and exerts a dominant-negative effect on its paralog, CCDC90B. We show that MCUR1 and CCDC90B form a hetero-oligomeric complex whose stability depends on MCUR1. Deletion of MCUR1 and CCDC90B in the fission yeast Schizosaccharomyces pombe, which lacks the MCUC, impairs lipid and amino acid metabolism and causes nitrogen source-dependent growth defects that are rescued by expression of human MCUR1. Patient serum metabolomics confirms an imbalance in the amino acid pool, while MCUR1 deficiency in patient-derived skin fibroblasts upregulates autophagy, perturbs non-essential amino acid metabolism, and limits biosynthetic capacity, resulting in delayed proliferation and migration. These findings redefine the MCUR1-CCDC90B coiled-coil complex as a transmembrane scaffold critical for the integrity of mitochondrial protein complexes and the maintenance of metabolic homeostasis, suggesting a potential link between MCUR1 deficiency and human neurometabolic disease.

cell biology↗

bioRxiv2024

A cautionary tale on the cost-effectiveness of collaborative AI in real-world medical applications.

Federated learning (FL) has gained wide popularity as a collaborative learning paradigm enabling trustworthy AI in sensitive healthcare applications. Never-theless, the practical implementation of FL presents technical and organizational challenges, as it generally requires complex communication infrastructures. In this context, consensus-based learning (CBL) may represent a promising collaborative learning alternative, thanks to the ability of combining local knowledge into a federated decision system, while potentially reducing deployment over-head. In this work we propose an extensive benchmark of the accuracy and cost-effectiveness of a panel of FL and CBL methods in a wide range of collaborative medical data analysis scenarios. Our results reveal that CBL is a cost-effective alternative to FL, providing comparable accuracy and significantly reducing training and communication costs. This study opens a novel perspective on the deployment of collaborative AI in real-world applications, whereas the adoption of cost-effective methods is instrumental to achieve sustainability and democratisation of AI by alleviating the need for extensive computational resources.

bioinformatics↗

bioRxiv2022

Weakly supervised Unet: an image classifier which learns to explain itself

BackgroundExplainability is a major limitation of current convolutional neural network (CNN) image classifiers. A CNN is required which supports its image-level prediction with a voxel-level segmentation. MethodsA weakly-supervised Unet architecture (WSUnet) is proposed to model voxel classes, by training with image-level supervision. WSUnet computes the image-level class prediction from the maximal voxel class prediction. Thus, voxel-level predictions provide a causally verifiable saliency map for the image-level decision. WSUnet is applied to explainable lung cancer detection in CT images. For comparison, current model explanation approaches are also applied to a standard CNN. Methods are compared using voxel-level discrimination metrics and a clinician preference survey. ResultsIn test data from two external institutions, WSUnet localised the tumour precisely at voxel-level (Precision: 0.93 [0.93-0.94]), achieving superior voxel-level discrimination to the best comparator (AUPR: 0.55 [0.54-0.55] vs. 0.36 [0.35-0.36]). Clinicians preferred WSUnet predictions in most test instances (Clinician Preference Rate: 0.72 [0.68-0.77]). ConclusionsWSUnet is a simple extension of the Unet, which facilitates voxel-level modelling from image-level labels. As WSUnet supports its image-level prediction with a causative voxel-level segmentation, it functions as a self-explaining image classifier. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=194 SRC="FIGDIR/small/507144v1_ufig1.gif" ALT="Figure 1"> View larger version (35K): org.highwire.dtl.DTLVardef@52e3ccorg.highwire.dtl.DTLVardef@1e981f0org.highwire.dtl.DTLVardef@151f31eorg.highwire.dtl.DTLVardef@13046a2_HPS_FORMAT_FIGEXP M_FIG Graphical Abstract The weakly-supervised Unet converts voxel-level predictions to image-level predictions using a global max-pooling layer. Thus, loss is computed at image-level. Following training with image-level labels, voxel-level predictions are extracted from the voxel-level output layer. C_FIG FundingAuthors acknowledge funding support from the UK Research & Innovation London Medical Imaging and Artificial Intelligence Centre; Wellcome/Engineering and Physical Sciences Research Council Centre for Medical Engineering at Kings College London [WT 203148/Z/16/Z]; National Institute for Health Research Biomedical Research Centre at Guys & St Thomas Hospitals and Kings College London; National Institute for Health Research Biomedical Research Centre at Guys & St Thomas Hospitals and Kings College London; Cancer Research UK National Cancer Imaging Translational Accelerator [C1519/A28682]. For the purpose of open access, authors have applied a CC BY public copyright licence to any Author Accepted Manuscript version arising from this submission. HIGHLIGHTSO_LIWSUnet is a weakly supervised Unet architecture which can learn semantic segmentation from data labelled only at image-level. C_LIO_LIWSUnet is a convolutional neural network image classifier which provides a causally verifiable voxel-level explanation to support its image-level prediction. C_LIO_LIIn application to explainable lung cancer detection, WSUnets voxel-level output localises tumours precisely, outperforming current model explanation methods. C_LIO_LIWSUnet is a simple extension of the standard Unet architecture, requiring only the addition of a global max-pooling layer to the output. C_LI

bioinformatics↗

bioRxiv2020

Discovery of EMRE in fungi resolves the true evolutionary history of the mitochondrial calciumuniporter

Mitochondrial calcium (mt-Ca2+) uptake is central for the regulation of numerous cellular processes in eukaryotes1. This occurs through a highly selective Ca2+ uniporter located at the inner mitochondrial membrane and driven by the membrane potential2-4. While the physiological role of the uniporter was extensively studied for decades, its genetic identity was only recently determined, with MCU5,6, MICU17 and EMRE8 constituting pore-forming and regulatory subunits. Preliminary evolutionary analyses suggested an ancient eukaryotic origin of mt-Ca2+ uptake, but also pinpointed inconsistent phylogenetic distributions of MCU, MICU1, and EMRE within fungi, where homologs of MCU were present in the absence of the supposedly essential regulators, MICU1 and EMRE9,10. Here, we perform the most comprehensive phylogenomic analysis of the mt-Ca2+ uptake system and trace its evolution across 1,156 fully-sequenced eukaryotes. In contrast to earlier assumptions9-11 we find compelling evidence that previously identified animal and fungal MCUs, the targets of several structural and functional efforts11-16, represent two distinct paralogous subfamilies originating from an ancestral duplication. We further uncover a complete "animal-like" uniporter complex within chytrid fungi, including bona-fide orthologs of MCU, MICU1, and EMRE. This first identification of EMRE outside Holozoa (animals and their unicellular relatives) and its strong coevolution with "animal-like" MICU1 and MCU indicates that these three components formed the core of the ancestral opisthokont uniporter. We confirm this finding experimentally, by showing that chytrid EMRE orthologs in combination with either human or "animal-like" MCUs, but not with "fungal-specific" MCUs, can reconstitute mt-Ca2+ uptake in vivo in the yeast Saccharomyces cerevisiae. Hence, we here solve a purported evolutionary paradox: the presence of MCU homologs in fungal species devoid of other uniporter components and with no detectable mt-Ca2+ uptake. Altogether, our study clarifies the evolution of the mt-Ca2+ uniporter and identifies new important targets for comparative structural and functional studies.

evolutionary biology↗

bioRxiv2018

MICU1 modulates MCU ion selectivity and tolerance to manganese stress

The mitochondrial calcium uniporter is a highly selective ion channel composed of species-and tissue-specific structural and regulatory subunits. However, the contribution of each component to uniporter-mediated activity still remains unclear. Here, we employ an evolutionary and synthetic biology approach to investigate the functional inter-dependence between the pore-forming subunit MCU and the EF-hand protein MICU1. Using phylogenetic profiling and genetic complementation analyses, we show that MCU and MICU1 constitute the minimal eukaryotic unit of the uniporter, pointing towards a strong selective pressure behind their co-occurrence. Heterologous reconstitution of MCU-mediated and MICU1-gated mitochondrial calcium entry in vivo in yeast cells demonstrates that MICU1 per se is essential to protect yeast from MCU-dependent manganese cytotoxicity. Accordingly, MICU1 deletion significantly sensitizes human HEK-293 cells to manganese-induced stress. Our study identifies a critical role of MICU1 in the regulation of MCU ion selectivity, with potential implications for patients with MICU1 deficiency.

cell biology↗