Featured
-
-
Article
| Open AccessA framework for evaluating clinical artificial intelligence systems without ground-truth annotations
Estimating the performance of clinical AI systems on data in the wild is complicated by distribution shift and the absence of ground-truth annotations. Here, we introduce SUDO, a framework for more reliably evaluating AI systems on data in the wild.
- Dani Kiyasseh
- , Aaron Cohen
- & Nicholas Altieri
-
Article
| Open AccessPrediction of plasma ctDNA fraction and prognostic implications of liquid biopsy in advanced prostate cancer
Metastatic castration-resistant prostate cancer is a highly aggressive disease, with a variable response to treatment. Here, the authors validate ctDNA fraction as a poor prognostic factor and develop a model to predict whether patients harbor sufficient ctDNA for informative blood-based genotyping.
- Nicolette M. Fonseca
- , Corinne Maurice-Dror
- & Alexander W. Wyatt
-
Article
| Open AccessFunctional regulation of aquaporin dynamics by lipid bilayer composition
Membrane proteins depend on their lipid environments. Using aquaporin as a model, the authors show that the choice of lipid bilayer fundamentally affects membrane protein structure, thermodynamics, kinetic, and function, even to the point of lipid-based inhibition.
- Anh T. P. Nguyen
- , Austin T. Weigle
- & Diwakar Shukla
-
Article
| Open AccessRare disease research workflow using multilayer networks elucidates the molecular determinants of severity in Congenital Myasthenic Syndromes
Congenital myasthenic syndromes are rare inherited neuromuscular disorders. Here, the authors attempt to explain diverse disease severity seen in 20 patients with shared CHRNE gene mutations with a multilayer network analysis that identifies individual-level impairments at the neuromuscular junction.
- Iker Núñez-Carpintero
- , Maria Rigau
- & Alfonso Valencia
-
Article
| Open AccessAutomatic data-driven design and 3D printing of custom ocular prostheses
Manual processes to produce ocular prostheses are time-consuming and yield varying quality. Here, authors present an automatic digital end-to-end process for custom ocular prostheses. It creates shape and appearance from image data of an OCT device and produces them using a full-colour 3D printer.
- Johann Reinhard
- , Philipp Urban
- & Mandeep S. Sagoo
-
Article
| Open AccessA release of local subunit conformational heterogeneity underlies gating in a muscle nicotinic acetylcholine receptor
Authors show that agonist binding to the muscle acetylcholine receptor releases local conformational heterogeneity transitioning all subunits into a symmetric open state. A release of conformational heterogeneity underlies allosteric communication.
- Mackenzie J. Thompson
- , Farid Mansoub Bekarkhanechi
- & John E. Baenziger
-
Article
| Open AccessStatistical method scDEED for detecting dubious 2D single-cell embeddings and optimizing t-SNE and UMAP hyperparameters
2D visualisation of single-cell data is highly impacted by the hyperparameter setting of the 2D embedding method, such as t-SNE and UMAP. Here, authors develop a statistical method scDEED to detect dubious cell embeddings and optimise the hyperparameter setting for trustworthy visualisation.
- Lucy Xia
- , Christy Lee
- & Jingyi Jessica Li
-
Article
| Open AccessA distinct class of pan-cancer susceptibility genes revealed by an alternative polyadenylation transcriptome-wide association study
Alternative polyadenylation (APA) can play a key role in cancer initiation and progression. Here, the authors conducted a comprehensive pan-cancer APA TWAS analysis and discovered a distinct class of APA-mediated cancer susceptibility genes across 22 cancer types.
- Hui Chen
- , Zeyang Wang
- & Lei Li
-
Article
| Open AccessSEMORE: SEgmentation and MORphological fingErprinting by machine learning automates super-resolution data analysis
There is a lack of universal tools to analyse protein assemblies and quantify underlying structures in single-molecule localization microscopy. Here, the authors present SEMORE, a semi-automatic machine learning framework for system- and input-dependent analysis of super-resolution data.
- Steen W. B. Bender
- , Marcus W. Dreisler
- & Nikos S. Hatzakis
-
Article
| Open AccessTransfer learning with graph neural networks for improved molecular property prediction in the multi-fidelity setting
Modern molecular discovery processes generate millions of measurements at different quality levels. Here, the authors develop a new deep learning method for transfer learning from low-cost and abundant data to enhance the efficiency of drug discovery.
- David Buterez
- , Jon Paul Janet
- & Pietro Lió
-
Article
| Open AccessRapid deep learning-assisted predictive diagnostics for point-of-care testing
A key aim in the development of diagnostic assays is improving diagnostic speed while maintaining sensitivity. Here the authors report an approach for the rapid and accurate analysis of lateral flow tests, which integrates time-series deep learning and AI verification, achieving a diagnostic time of 1-2 minutes.
- Seungmin Lee
- , Jeong Soo Park
- & Jeong Hoon Lee
-
Article
| Open AccessMetabolomic machine learning predictor for diagnosis and prognosis of gastric cancer
Gastric cancer detection by endoscopy is intrusive and time-consuming, and early detection is key to improving survival. Here, the authors propose a metabolite-based model to enable early detection.
- Yangzi Chen
- , Bohong Wang
- & Zeping Hu
-
Article
| Open AccessComplex regulatory networks influence pluripotent cell state transitions in human iPSCs
Stem cells exist in vitro in a spectrum of interconvertible pluripotent states. Here, authors show that pluripotency and self-renewal processes have a high level of regulatory complexity and suggest that genetic factors contribute to cell state transitions in human iPSC lines.
- Timothy D. Arthur
- , Jennifer P. Nguyen
- & Kelly A. Frazer
-
Article
| Open AccessscCASE: accurate and interpretable enhancement for single-cell chromatin accessibility sequencing data
Single-cell chromatin accessibility sequencing (scCAS) data suffers from high sparsity and dimensionality. Here, authors propose an accurate and interpretable computational framework for enhancing scCAS data that considers cell-to-cell similarity.
- Songming Tang
- , Xuejian Cui
- & Shengquan Chen
-
Article
| Open AccessProtein design using structure-based residue preferences
Recent protein design methods rely on large neural networks, yet it is unclear which dependencies are critical for determining function. Here, authors show that learning the per residue mutation preferences, without considering interactions, enables design of functional and diverse protein variants.
- David Ding
- , Ada Y. Shaw
- & Debora S. Marks
-
Article
| Open AccessDesign of target specific peptide inhibitors using generative deep learning and molecular dynamics simulations
Here the authors report a computational approach which integrates deep learning and structural modelling to design target-specific peptides. They apply this to β-catenin and NF-κB essential modulator, resulting in improved binding, highlighting the efficacy of this strategy.
- Sijie Chen
- , Tong Lin
- & Xiaolin Cheng
-
Article
| Open AccessAccurate global and local 3D alignment of cryo-EM density maps using local spatial structural features
Density map alignment is a fundamental step in Cryo-EM data postprocessing. Here, authors propose an accurate global and local density map alignment method using local density features.
- Bintao He
- , Fa Zhang
- & Renmin Han
-
Article
| Open AccessLearning representations for image-based profiling of perturbations
Assessing cell phenotypes in image-based assays requires solid computational methods for transforming images into quantitative data. Here, the authors present a strategy for learning representations of treatment effects from high-throughput imaging, following a causal interpretation.
- Nikita Moshkov
- , Michael Bornholdt
- & Juan C. Caicedo
-
Article
| Open AccessEfficient encoding of large antigenic spaces by epitope prioritization with Dolphyn
Profiling antibody responses to vast antigenic spaces has been challenging using programmable phage display (PhIP-Seq). Here, authors develop a methodology for compressing large proteomic spaces and have discovered human antibodies targeting gut bacteria-infecting phages.
- Anna-Maria Liebhoff
- , Thiagarajan Venkataraman
- & H. Benjamin Larman
-
Article
| Open AccessMulti-omics analysis in human retina uncovers ultraconserved cis-regulatory elements at rare eye disease loci
Ultraconserved non-coding elements (UCNEs) can regulate developmental gene expression. Retinal multi-omics data integration revealed UCNEs to be candidate cis-regulatory elements during retinal development, which may be implicated in rare eye diseases.
- Victor Lopez Soriano
- , Alfredo Dueñas Rey
- & Elfride De Baere
-
Article
| Open AccessLarge language models streamline automated machine learning for clinical studies
A knowledge gap persists between machine learning developers and clinicians. Here, the authors show that the Advanced Data Analysis extension of ChatGPT could bridge this gap and simplify complex data analyses, making them more accessible to clinicians.
- Soroosh Tayebi Arasteh
- , Tianyu Han
- & Sven Nebelung
-
Article
| Open AccessIterative design of training data to control intricate enzymatic reaction networks
Kinetic modeling of in vitro enzymatic reaction networks (ERNs) is severely hampered by the lack of training data. Here, authors introduce a methodology that combines an active learning-like approach and flow chemistry to create optimized datasets for an intricate ERN.
- Bob van Sluijs
- , Tao Zhou
- & Wilhelm T. S. Huck
-
Article
| Open AccessImputation of plasma lipid species to facilitate integration of lipidomic datasets
Advancements in plasma lipidomic profiling increase specificity of measurements but pose challenges in aligning datasets created at different times or platforms. Here the authors present a predictive framework for harmonising such datasets with different levels of granularity in their lipid measurements.
- Aleksandar Dakic
- , Jingqin Wu
- & Peter J. Meikle
-
Article
| Open AccessUnsupervised classification of brain-wide axons reveals the presubiculum neuronal projection blueprint
The classification of different types of neurons has been a long-standing challenge in neuroscience. Here, the authors present a strategy to quantify all statistically distinct axonal patterns from a brain region based on their anatomical targeting, with this projection-driven neuron classification informing the functional architecture of the circuit.
- Diek W. Wheeler
- , Shaina Banduri
- & Giorgio A. Ascoli
-
Article
| Open AccessAn agricultural digital twin for mandarins demonstrates the potential for individualized agriculture
A digital twin represents a real world object using available data. Here, the authors develop a digital twin for mandaring orchards in Jeju island showing the value of individualized agriculture to predict fruit quality at tree level.
- Steven Kim
- & Seong Heo
-
Article
| Open AccessA mutational atlas for Parkin proteostasis
Gene variants can affect folding and stability of the encoded protein. Here, the authors apply deep mutational scanning to provide genotype-phenotype information for 99% of the possible PRKN variants and reveal mechanistic details on how some variants cause loss-of-function and Parkinsons disease.
- Lene Clausen
- , Vasileios Voutsinos
- & Rasmus Hartmann-Petersen
-
Article
| Open AccessRecurrent evolutionary switches of mitochondrial cytochrome c maturation systems in Archaeplastida
Cytochrome c maturation (CCM) is the process of covalent attachment of a heme group to the conserved cysteines to form the holocytochrome. Here, the authors report that the non-adaptive convergent evolution at the pathway level leads to mosaic distribution of CCM systems I and III among Archaeplastida species.
- Huang Li
- , Soujanya Akella
- & Jeffrey P. Mower
-
Article
| Open AccessIdentification of HDV-like theta ribozymes involved in tRNA-based recoding of gut bacteriophages
The diverse functional roles of ribozymes (RNAs with enzymatic activity) continue to be uncovered. Here, the authors identify and characterize a subgroup of minimal hepatitis delta virus (HDV)-like ribozymes – termed Theta ribozymes -, which they show process viral tRNA transcripts, and appearing crucial for lytic gene expression in recoded phages.
- Kasimir Kienbeck
- , Lukas Malfertheiner
- & Roland K. O. Sigel
-
Article
| Open AccessPhage-plasmids promote recombination and emergence of phages and plasmids
Phage-plasmids are mobile genetic elements that transfer horizontally between bacterial cells as viruses, and vertically within bacterial lineages as plasmids. Here, Pfeifer & Rocha show that phage-plasmids can mediate gene transfer across mobile elements within their hosts, and can act as intermediates in the conversion of one type of element into another.
- Eugen Pfeifer
- & Eduardo P. C. Rocha
-
Article
| Open AccessMulti-dimensional scaling techniques unveiled gain1q&loss13q co-occurrence in Multiple Myeloma patients with specific genomic, transcriptional and adverse clinical features
The characterisation of the molecular features of multiple myeloma (MM) remains challenging. Here, the authors identify a subset of MM patients with a dismal clinical outcome, harbouring both chromosomes 1q CN gain and 13 CN loss and overexpressing CCND2.
- Carolina Terragna
- , Andrea Poletti
- & Michele Cavo
-
Article
| Open AccessMachine learning-based extrachromosomal DNA identification in large-scale cohorts reveals its clinical implications in cancer
‘Extrachromosomal DNA has been previously linked to tumour progression and heterogeneity, but its potential as a cancer biomarker has not been fully explored. Here, the authors develop a computational framework to refine genomic subtypes and predict response to immunotherapy in gastrointestinal cancer.
- Shixiang Wang
- , Chen-Yi Wu
- & Qi Zhao
-
Article
| Open AccessA signal processing and deep learning framework for methylation detection using Oxford Nanopore sequencing
The authors present DeepMod2, a deep-learning based computational method that allows fast and accurate detection of DNA methylation and epihaplotypes from Oxford Nanopore sequencing data.
- Mian Umair Ahsan
- , Anagha Gouru
- & Kai Wang
-
Article
| Open AccessA genome and gene catalog of the aquatic microbiomes of the Tibetan Plateau
The Tibetan Plateau is the largest plateau in the world and hosts a variety of aquatic ecosystems. Here, the authors present a gene and genome catalogue of Tibetan Plateau aquatic microbiomes, greatly expanding known taxonomic and functional diversity for the region and giving insights into its microbial biogeography.
- Mingyue Cheng
- , Shuai Luo
- & Kang Ning
-
Comment
| Open AccessFudging the volcano-plot without dredging the data
Selecting omic biomarkers using both their effect size and their differential status significance (i.e., selecting the “volcano-plot outer spray”) has long been equally biologically relevant and statistically troublesome. However, recent proposals are paving the way to resolving this dilemma.
- Thomas Burger
-
Article
| Open AccessTFvelo: gene regulation inspired RNA velocity estimation
Most RNA velocity models extract dynamics from the phase delay between unspliced and spliced mRNA for each gene. Here, authors propose TFvelo, broadening RNA velocity beyond splicing information to include gene regulation. TFvelo accurately models genes dynamics and infers cell pseudo-time from RNA abundance data.
- Jiachen Li
- , Xiaoyong Pan
- & Hong-Bin Shen
-
Article
| Open AccessPheWAS-based clustering of Mendelian Randomisation instruments reveals distinct mechanism-specific causal effects between obesity and educational attainment
Mendelian Randomisation estimates causal effects between risk factors and complex outcomes using genetic variants as instrumental variables, however it can be affected by certain biases. To alleviate these biases the authors propose an approach based on clustering genetic instruments according to the types of trait they are associated with, and apply this method to revisit the surprisingly large apparent causal effect of body mass index on educational attainment.
- Liza Darrous
- , Gibran Hemani
- & Zoltán Kutalik
-
Article
| Open AccessHigh resolution spatial profiling of kidney injury and repair using RNA hybridization-based in situ sequencing
Advancements in spatial transcriptomics technologies have enabled the analysis of gene expression at cellular resolution in situ. The authors applied direct RNA hybridization-based in situ sequencing (dRNA HybISS) and developed a computational tool, CellScopes, to study gene expression in mouse kidneys, identifying cellular changes and interactions during injury and repair.
- Haojia Wu
- , Eryn E. Dixon
- & Benjamin D. Humphreys
-
Article
| Open AccessA deep-learning-based framework for identifying and localizing multiple abnormalities and assessing cardiomegaly in chest X-ray
Accurate localization of abnormalities is crucial in the interpretation of chest X-rays. Here the authors present a deep learning framework for simultaneous localization of 14 thoracic abnormalities and calculation of cardiothoracic ratio, based on large X-ray dataset with bounding boxes created via a human-in-the-loop approach.
- Weijie Fan
- , Yi Yang
- & Dong Zhang
-
Article
| Open AccessGenomic epidemiology reveals geographical clustering of multidrug-resistant Escherichia coli ST131 associated with bacteraemia in Wales
Escherichia coli ST131 is a globally dominant multidrug resistant clone associated with high rates of recurring urinary tract infections. In this genomic epidemiology study, the authors describe the evolution, population structure, and antimicrobial resistance in 142 E. coli ST131 samples from Wales, UK.
- Rhys T. White
- , Matthew J. Bull
- & Scott A. Beatson
-
Article
| Open AccessA general computational design strategy for stabilizing viral class I fusion proteins
The authors present a generalisable computational approach to stabilize class I fusion proteins in the prefusion state. The method was used to stabilize the fusion proteins of RSV, hMPV, and SARS-CoV-2 viruses, with the designs structurally validated and RSV F protein assessed in a neutralization assay.
- Karen J. Gonzalez
- , Jiachen Huang
- & Eva-Maria Strauch
-
Article
| Open AccessPredicting proximal tubule failed repair drivers through regularized regression analysis of single cell multiomic sequencing
A profibrotic, proinflammatory kidney cell population has been identified as a driver of chronic kidney disease. Here, authors generate a human kidney single cell multiomic dataset and apply a regularised regression approach to identify transcription factors underpinning this cell population.
- Nicolas Ledru
- , Parker C. Wilson
- & Benjamin D. Humphreys
-
Article
| Open AccessRegression-based Deep-Learning predicts molecular biomarkers from pathology slides
Cancer biomarkers are often continuous measurements, which poses challenges for their prediction using classification-based deep learning. Here, the authors develop a regression-based deep learning method to predict continuous biomarkers - such as the homologous repair deficiency score - from cancer histopathology images.
- Omar S. M. El Nahhas
- , Chiara M. L. Loeffler
- & Jakob Nikolas Kather
-
Article
| Open AccessMatrin3 mediates differentiation through stabilizing chromatin loop-domain interactions and YY1 mediated enhancer-promoter interactions
Alterations in proteins within nuclear compartments often lead to changes in chromosomal architecture. Here, using acute targeted protein degradation, the authors reveal that the nuclear complex protein Matrin3 directly mediates differentiation through stabilizing chromatin loop domain interactions.
- Tianxin Liu
- , Qian Zhu
- & Stuart H. Orkin
-
Article
| Open AccessA method to estimate the contribution of rare coding variants to complex trait heritability
The contribution of rare variants to complex traits has not been well studied. Here, the authors present RARity, a method to assess rare variant heritability without assuming a particular genetic architecture and enabling both gene-level and exome-wide heritability estimation of continuous traits.
- Nazia Pathan
- , Wei Q. Deng
- & Guillaume Paré
-
Article
| Open AccessPredicting DNA structure using a deep learning method
In this work, the authors report a deep learning method, Deep DNAshape, to predict the influence of flanking regions on three-dimensional DNA structure and in structural readout mechanisms of protein-DNA binding.
- Jinsen Li
- , Tsu-Pei Chiu
- & Remo Rohs
-
Review Article
| Open AccessThe genetic basis of autoimmunity seen through the lens of T cell functional traits
Genetic risk variants for autoimmune diseases are largely enriched in T cell-specific regulatory regions. In this review, Raychaudhuri and colleagues summarise the findings of recent studies evaluating the genetic regulation of T cell molecular and functional traits in these diseases.
- Kaitlyn A. Lagattuta
- , Hannah L. Park
- & Soumya Raychaudhuri
-
Article
| Open AccessAntiviral fibrils of self-assembled peptides with tunable compositions
In this work, the authors report the use of a computationally and rationally designed self-assembling peptide that has robust antiviral capability with demonstrated specificity in binding to SARS-CoV-2 and inhibition of viral entry into human cells.
- Joseph Dodd-o
- , Abhishek Roy
- & Vivek Kumar
-
Article
| Open AccessA multicenter clinical AI system study for detection and diagnosis of focal liver lesions
Early detection and accurate diagnosis of focal liver lesions are crucial for effective treatment and prognosis. Here, the authors present a fully automated diagnostic system that leverages multi-phase CT scans and clinical features, for diagnosing liver lesions.
- Hanning Ying
- , Xiaoqing Liu
- & Xiujun Cai
-
Article
| Open AccessOverlay databank unlocks data-driven analyses of biomolecules for all
In this work, the authors report NMR lipids Databank to promote decentralised sharing of biomolecular molecular dynamics (MD) simulation data with an overlay design. Programmatic access enables analyses of rare phenomena and advances the training of machine learning models.
- Anne M. Kiirikki
- , Hanne S. Antila
- & O. H. Samuli Ollila
Browse broader subjects
Browse narrower subjects
- Biochemical reaction networks
- Cellular signalling networks
- Classification and taxonomy
- Communication and replication
- Computational models
- Computational neuroscience
- Computational platforms and environments
- Data acquisition
- Data integration
- Data mining
- Data processing
- Data publication and archiving
- Databases
- Functional clustering
- Gene ontology
- Gene regulatory networks
- Genome informatics
- Hardware and infrastructure
- High-throughput screening
- Image processing
- Literature mining
- Machine learning
- Microarrays
- Network topology
- Phylogeny
- Power law
- Predictive medicine
- Probabilistic data networks
- Programming language
- Protein analysis
- Protein design
- Protein folding
- Protein function predictions
- Protein structure predictions
- Proteome informatics
- Quality control
- Scale invariance
- Sequence annotation
- Software
- Standards
- Statistical methods
- Virtual drug screening