Data Brief. 2023 Jul 20;49:109433. doi: 10.1016/j.dib.2023.109433. eCollection 2023 Aug.

ABSTRACT

The immunoprotective functions of microglia in the brain are mediated by the inflammatory M1 phenotype. This phenotype is challenged by anti-inflammatory cytokines which polarize the microglia cells to an immunosuppressive M2 phenotype, a trait that is often exploited by cancer cells to evade immune recognition and promote tumor growth. Investigating the molecular determinants of this behavior is crucial for advancing the understanding of the mechanisms that cancer cells use to escape immune attack. In this article, we describe liquid chromatography (LC)-mass spectrometry (MS)/proteomic data acquired with an EASY-nanoLC 1200-Q ExactiveTM OrbitrapTM mass spectrometer that reflect the response of human microglia cells (HMC3) to stimulation with potential cancer-released anti-inflammatory cytokines known to be key players in promoting tumorigenesis in the brain (IL-4, IL-13, IL-10, TGFB and MCP-1). The MS files were processed with the Proteome Discoverer v.2.4 software package. The cell culture conditions, the sample preparation protocols, the MS acquisition parameters, and the data processing approach are described in detail. The RAW and processed MS files associated with this work were deposited in the PRIDE partner repository of the ProteomeXchange Consortium with the dataset identifiers PXD023163 and PXD023166, and the analyzed data in the Mendeley Data cloud-based repository with DOI 10.17632/fvhw2zwt5d.1. The biological interpretation of the data can be accessed in the research article "Systems-Level Proteomics Evaluation of Microglia Response to Tumor-Supportive Anti-inflammatory Cytokines" (Shreya Ahuja and Iulia M. Lazar, Frontiers in Immunology 2021 [1]). The proteome data described in this article will benefit researchers who are either interested in re-processing the data with alternative search engines and filtering criteria, and/or exploring the data in more depth to advance the understanding of cancer progression and the discovery of novel biomarkers or drug targets.

PMID:37554993 | PMC:PMC10404857 | DOI:10.1016/j.dib.2023.109433

iScience. 2023 Jul 11;26(8):107300. doi: 10.1016/j.isci.2023.107300. eCollection 2023 Aug 18.

ABSTRACT

METTL3 is the major writer of N6-Methyladenosine (m6A) and has been associated with controversial roles in cancer. This is best illustrated in urothelial carcinoma of the bladder (UCB), where METTL3 was described to have both oncogenic and tumor-suppressive functions. Here, we reinvestigated the role of METTL3 in UCB. METTL3 knockout reduced the oncogenic phenotype and m6A levels of UCB cell lines. However, complete depletion of METTL3/m6A was not achieved due to selection of cells expressing alternative METTL3 isoforms. Systematic vulnerability and inhibitor response analyses suggested that uroepithelial cells depend on METTL3 for viability. Furthermore, expression and survival analyses of clinical data revealed a complex role for METTL3 in UCB, with decreased m6A mRNA levels in UCB tumors. Our results suggest that METTL3 expression may be a suitable diagnostic UCB biomarker, as the enzyme promotes UCB formation. However, the suitability of the enzyme as a therapeutic target should be evaluated carefully.

PMID:37554463 | PMC:PMC10405067 | DOI:10.1016/j.isci.2023.107300

iScience. 2023 Jul 10;26(8):107302. doi: 10.1016/j.isci.2023.107302. eCollection 2023 Aug 18.

ABSTRACT

This study investigates the role of integrin β4 (ITGB4) and stemness-associated factor SOX2 in platinum resistance in lung squamous cell carcinoma (LUSC). The expression of SOX2 and ITGB4 is found to be high in all LUSC subtypes, but the impact of ITGB4 expression on overall patient survival varies by subtype. Cancer stem cells (CSCs) isolated from LUSC patients were found to be resistant to cisplatin, but knocking down ITGB4 or SOX2 sensitized them to cisplatin. Carfilzomib (CFZ) synergized with cisplatin and suppressed CSC growth by inhibiting ITGB4 and SOX2 expression. Additionally, CFZ was found to inhibit SOX2 expression epigenetically by inhibiting histone acetylation at the SOX2 promoter site. CFZ also suppressed the growth of SOX2-dependent small cell lung cancer cells in vitro and in vivo. The study highlights the unique function of CFZ as a transcriptional suppressor of SOX2, independent of its proteasome inhibitory function.

PMID:37554452 | PMC:PMC10405066 | DOI:10.1016/j.isci.2023.107302

Genome Biol. 2023 Aug 8;24(1):183. doi: 10.1186/s13059-023-03028-2.

ABSTRACT

BACKGROUND: Cumulative sequencing efforts have yielded enough genomes to construct pangenomes for dozens of bacterial species and elucidate intraspecies gene conservation. Given the diversity of organisms for which this is achievable, similar analyses for ancestral species are feasible through the integration of pangenomics and phylogenetics, promising deeper insights into the nature of ancient life.

RESULTS: We construct pangenomes for 183 bacterial species from 54,085 genomes and identify their core genomes using a novel statistical model to estimate genome-specific error rates and underlying gene frequencies. The core genomes are then integrated into a phylogenetic tree to reconstruct the core genome of the last bacterial common ancestor (LBCA), yielding three main results: First, the gene content of modern and ancestral core genomes are diverse at the level of individual genes but are similarly distributed by functional category and share several poorly characterized genes. Second, the LBCA core genome is distinct from any individual modern core genome but has many fundamental biological systems intact, especially those involving translation machinery and biosynthetic pathways to all major nucleotides and amino acids. Third, despite this metabolic versatility, the LBCA core genome likely requires additional non-core genes for viability, based on comparisons with the minimal organism, JCVI-Syn3A.

CONCLUSIONS: These results suggest that many cellular systems commonly conserved in modern bacteria were not just present in ancient bacteria but were nearly immutable with respect to short-term intraspecies variation. Extending this analysis to other domains of life will likely provide similar insights into more distant ancestral species.

PMID:37553643 | DOI:10.1186/s13059-023-03028-2

Sci Rep. 2023 Aug 8;13(1):12830. doi: 10.1038/s41598-023-40096-7.

ABSTRACT

Grouper culture has been expanding in Malaysia due to the huge demand locally and globally. However, due to infectious diseases such as vibriosis, the fish mortality rate increased, which has affected the production of grouper. Therefore, this study focuses on the metabolic profiling of surviving infected grouper fed with different formulations of fatty acid diets that acted as immunostimulants for the fish to achieve desirable growth and health performance. After a six-week feeding trial and one-week post-bacterial challenge, the surviving infected grouper was sampled for GC-MS analysis. For metabolite extraction, a methanol/chloroform/water (2:2:1.8) extraction method was applied to the immune organs (spleen and liver) of surviving infected grouper. The distribution patterns of metabolites between experimental groups were then analyzed using a metabolomics platform. A total of 50 and 81 metabolites were putatively identified from the spleen and liver samples, respectively. Our further analysis identified glycine, serine, and threonine metabolism, and alanine, aspartate and glutamate metabolism had the most impacted pathways, respectively, in spleen and liver samples from surviving infected grouper. The metabolites that were highly abundant in the spleen found in these pathways were glycine (20.9%), l-threonine (1.0%) and l-serine (0.8%). Meanwhile, in the liver l-glutamine (1.8%) and aspartic acid (0.6%) were found to be highly abundant. Interestingly, among the fish diet groups, grouper fed with oleic acid diet produced more metabolites with a higher percent area compared to the control diets. The results obtained from this study elucidate the use of oleic acid as an immunostimulant in fish feed formulation affects more various immune-related metabolites than other formulated feed diets for vibriosis infected grouper.

PMID:37553472 | DOI:10.1038/s41598-023-40096-7

Nat Commun. 2023 Aug 8;14(1):4768. doi: 10.1038/s41467-023-40360-4.

ABSTRACT

Metagenomic or metabarcoding data are often used to predict microbial interactions in complex communities, but these predictions are rarely explored experimentally. Here, we use an organism abundance correlation network to investigate factors that control community organization in mine tailings-derived laboratory microbial consortia grown under dozens of conditions. The network is overlaid with metagenomic information about functional capacities to generate testable hypotheses. We develop a metric to predict the importance of each node within its local network environments relative to correlated vitamin auxotrophs, and predict that a Variovorax species is a hub as an important source of thiamine. Quantification of thiamine during the growth of Variovorax in minimal media show high levels of thiamine production, up to 100 mg/L. A few of the correlated thiamine auxotrophs are predicted to produce pantothenate, which we show is required for growth of Variovorax, supporting that a subset of vitamin-dependent interactions are mutualistic. A Cryptococcus yeast produces the B-vitamin pantothenate, and co-culturing with Variovorax leads to a 90-130-fold fitness increase for both organisms. Our study demonstrates the predictive power of metagenome-informed, microbial consortia-based network analyses for identifying microbial interactions that underpin the structure and functioning of microbial communities.

PMID:37553333 | DOI:10.1038/s41467-023-40360-4

Nat Commun. 2023 Aug 8;14(1):4781. doi: 10.1038/s41467-023-40498-1.

ABSTRACT

Metabolic engineering of microalgae offers a promising solution for sustainable biofuel production, and rational design of engineering strategies can be improved by employing metabolic models that integrate enzyme turnover numbers. However, the coverage of turnover numbers for Chlamydomonas reinhardtii, a model eukaryotic microalga accessible to metabolic engineering, is 17-fold smaller compared to the heterotrophic cell factory Saccharomyces cerevisiae. Here we generate quantitative protein abundance data of Chlamydomonas covering 2337 to 3708 proteins in various growth conditions to estimate in vivo maximum apparent turnover numbers. Using constrained-based modeling we provide proxies for in vivo turnover numbers of 568 reactions, representing a 10-fold increase over the in vitro data for Chlamydomonas. Integration of the in vivo estimates instead of in vitro values in a metabolic model of Chlamydomonas improved the accuracy of enzyme usage predictions. Our results help in extending the knowledge on uncharacterized enzymes and improve biotechnological applications of Chlamydomonas.

PMID:37553325 | DOI:10.1038/s41467-023-40498-1

Discov Med. 2023 Aug;35(177):553-564. doi: 10.24976/Discov.Med.202335177.56.

ABSTRACT

PURPOSE: To evaluate the effects of various retinal neurotransmitters on temporal resolution, particularly, on the Critical Flicker Fusion Frequency (CFF), which has been previously applied in ophthalmic pathophysiologic research.

METHODS: A binocular physiologic electroretinogram was performed on adult mice. Animals in the control group were injected in the right eye with 1 μL of phosphate-buffered saline (PBS). Animals in the experimental group were injected in the left eye with 1 μL of PBS and in the right eye with 1 μL of PBS to which different molecules were added: 2-amino-4-phosphonobutyric acid (APB), Glutamate, γ-aminobutyric acid (GABA), 6,7-dinitroquinoxaline-2,3-dione (DNQX), Bicuculline, Glycine, and 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES). Initially, rod response was recorded and later the cone response.

RESULTS: APB suppressed the rod-driven, but not the cone-driven flicker response. The other agents severely affected the lower flickering frequency response amplitude, in particular, at 3 Hz. The threshold of CFF was lowered from 50 Hz to 40 Hz after applying APB, Glycine, and HEPES. GABA remarkably enhanced rod-driven and cone-driven flicker response at 3 Hz, whereas Glutamate and GABA/Glutamate only did in rod-driven flicker response.

CONCLUSIONS: Both ON and OFF visual pathways were implied in cone-driven response, but only the ON visual pathway appears to play a relevant role in rod-driven flicker response. Flicker response seems to be enhanced by horizontal cells both in rod-driven and cone-driven response. In addition, due to the greater sensitivity of the flicker at low frequencies, it is suggested that pathophysiological studies should be carried out at said frequencies.

PMID:37553309 | DOI:10.24976/Discov.Med.202335177.56

Ann Rev Mar Sci. 2023 Aug 8. doi: 10.1146/annurev-marine-022123-104345. Online ahead of print.

ABSTRACT

The holobiont concept (i.e., multiple living beings in close symbiosis with one another and functioning as a unit) is revolutionizing our understanding of biology, especially in marine systems. The earliest marine holobiont was likely a syntrophic partnership of at least two prokaryotic members. Since then, symbiosis has enabled marine organisms to conquer all ocean habitats through the formation of holobionts with a wide spectrum of complexities. However, most scientific inquiries have focused on isolated organisms and their adaptations to specific environments. In this review, we attempt to illustrate why a holobiont perspective-specifically, the study of how numerous organisms form a discrete ecological unit through symbiosis-will be a more impactful strategy to advance our understanding of the ecology and evolution of marine life. We argue that this approach is instrumental in addressing the threats to marine biodiversity posed by the current global environmental crisis. Expected final online publication date for the Annual Review of Marine Science, Volume 16 is January 2024. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

PMID:37552896 | DOI:10.1146/annurev-marine-022123-104345

Annu Rev Genet. 2023 Aug 8. doi: 10.1146/annurev-genet-080320-031523. Online ahead of print.

ABSTRACT

Transcription and replication both require large macromolecular complexes to act on a DNA template, yet these machineries cannot simultaneously act on the same DNA sequence. Conflicts between the replication and transcription machineries (transcription-replication conflicts, or TRCs) are widespread in both prokaryotes and eukaryotes and have the capacity to both cause DNA damage and compromise complete, faithful replication of the genome. This review will highlight recent studies investigating the genomic locations of TRCs and the mechanisms by which they may be prevented, mitigated, or resolved. We address work from both model organisms and mammalian systems but predominantly focus on multicellular eukaryotes owing to the additional complexities inherent in the coordination of replication and transcription in the context of cell type-specific gene expression and higher-order chromatin organization. Expected final online publication date for the Annual Review of Genetics, Volume 57 is November 2023. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

PMID:37552891 | DOI:10.1146/annurev-genet-080320-031523

Proc Natl Acad Sci U S A. 2023 Aug 15;120(33):e2301411120. doi: 10.1073/pnas.2301411120. Epub 2023 Aug 8.

ABSTRACT

The acquisition of novel sexually dimorphic traits poses an evolutionary puzzle: How do new traits arise and become sex-limited? Recently acquired color vision, sexually dimorphic in animals like primates and butterflies, presents a compelling model for understanding how traits become sex-biased. For example, some Heliconius butterflies uniquely possess UV (ultraviolet) color vision, which correlates with the expression of two differentially tuned UV-sensitive rhodopsins, UVRh1 and UVRh2. To discover how such traits become sexually dimorphic, we studied Heliconius charithonia, which exhibits female-specific UVRh1 expression. We demonstrate that females, but not males, discriminate different UV wavelengths. Through whole-genome shotgun sequencing and assembly of the H. charithonia genome, we discovered that UVRh1 is present on the W chromosome, making it obligately female-specific. By knocking out UVRh1, we show that UVRh1 protein expression is absent in mutant female eye tissue, as in wild-type male eyes. A PCR survey of UVRh1 sex-linkage across the genus shows that species with female-specific UVRh1 expression lack UVRh1 gDNA in males. Thus, acquisition of sex linkage is sufficient to achieve female-specific expression of UVRh1, though this does not preclude other mechanisms, like cis-regulatory evolution from also contributing. Moreover, both this event, and mutations leading to differential UV opsin sensitivity, occurred early in the history of Heliconius. These results suggest a path for acquiring sexual dimorphism distinct from existing mechanistic models. We propose a model where gene traffic to heterosomes (the W or the Y) genetically partitions a trait by sex before a phenotype shifts (spectral tuning of UV sensitivity).

PMID:37552755 | DOI:10.1073/pnas.2301411120

G3 (Bethesda). 2023 Aug 8:jkad177. doi: 10.1093/g3journal/jkad177. Online ahead of print.

ABSTRACT

There is increasing interest in the African spiny mouse (Acomys cahirinus) as a model organism because of its ability for regeneration of tissue after injury in skin, muscle, and internal organs such as the kidneys. A high-quality reference genome is needed to better understand these regenerative properties at the molecular level. Here, we present an improved reference genome for A. cahirinus generated from long Nanopore sequencing reads. We confirm the quality of our annotations using RNA sequencing data from four different tissues. Our genome is of higher contiguity and quality than previously reported genomes from this species and will facilitate ongoing efforts to better understand the regenerative properties of this organism.

PMID:37552705 | DOI:10.1093/g3journal/jkad177

Brief Bioinform. 2023 Aug 7:bbad274. doi: 10.1093/bib/bbad274. Online ahead of print.

ABSTRACT

Prediction of driver genes (tumor suppressors and oncogenes) is an essential step in understanding cancer development and discovering potential novel treatments. We recently proposed Moonlight as a bioinformatics framework to predict driver genes and analyze them in a system-biology-oriented manner based on -omics integration. Moonlight uses gene expression as a primary data source and combines it with patterns related to cancer hallmarks and regulatory networks to identify oncogenic mediators. Once the oncogenic mediators are identified, it is important to include extra levels of evidence, called mechanistic indicators, to identify driver genes and to link the observed gene expression changes to the underlying alteration that promotes them. Such a mechanistic indicator could be for example a mutation in the regulatory regions for the candidate gene. Here, we developed new functionalities and released Moonlight2 to provide the user with a mutation-based mechanistic indicator as a second layer of evidence. These functionalities analyze mutations in a cancer cohort to classify them into driver and passenger mutations. Those oncogenic mediators with at least one driver mutation are retained as the final set of driver genes. We applied Moonlight2 to the basal-like breast cancer subtype, lung adenocarcinoma and thyroid carcinoma using data from The Cancer Genome Atlas. For example, in basal-like breast cancer, we found four oncogenes (COPZ2, SF3B4, KRTCAP2 and POLR2J) and nine tumor suppressor genes (KIR2DL4, KIF26B, ARL15, ARHGAP25, EMCN, GMFG, TPK1, NR5A2 and TEK) containing a driver mutation in their promoter region, possibly explaining their deregulation. Moonlight2R is available at https://github.com/ELELAB/Moonlight2R.

PMID:37551622 | DOI:10.1093/bib/bbad274

BMC Genomics. 2023 Aug 7;24(1):443. doi: 10.1186/s12864-023-09506-y.

ABSTRACT

BACKGROUND: Morphological and traditional genetic studies of the young Pliocene genus Hyles have led to the understanding that despite its importance for taxonomy, phenotypic similarity of wing patterns does not correlate with phylogenetic relationship. To gain insights into various aspects of speciation in the Spurge Hawkmoth (Hyles euphorbiae), we assembled a chromosome-level genome and investigated some of its characteristics.

RESULTS: The genome of a male H. euphorbiae was sequenced using PacBio and Hi-C data, yielding a 504 Mb assembly (scaffold N50 of 18.2 Mb) with 99.9% of data represented by the 29 largest scaffolds forming the haploid chromosome set. Consistent with this, FISH analysis of the karyotype revealed n = 29 chromosomes and a WZ/ZZ (female/male) sex chromosome system. Estimates of chromosome length based on the karyotype image provided an additional quality metric of assembled chromosome size. Rescaffolding the published male H. vespertilio genome resulted in a high-quality assembly (651 Mb, scaffold N50 of 22 Mb) with 98% of sequence data in the 29 chromosomes. The larger genome size of H. vespertilio (average 1C DNA value of 562 Mb) was accompanied by a proportional increase in repeats from 45% in H. euphorbiae (measured as 472 Mb) to almost 55% in H. vespertilio. Several wing pattern genes were found on the same chromosomes in the two species, with varying amounts and positions of repetitive elements and inversions possibly corrupting their function.

CONCLUSIONS: Our two-fold comparative genomics approach revealed high gene synteny of the Hyles genomes to other Sphingidae and high correspondence to intact Merian elements, the ancestral linkage groups of Lepidoptera, with the exception of three simple fusion events. We propose a standardized approach for genome taxonomy using nucleotide homology via scaffold chaining as the primary tool combined with Oxford plots based on Merian elements to infer and visualize directionality of chromosomal rearrangements. The identification of wing pattern genes promises future understanding of the evolution of forewing patterns in the genus Hyles, although further sequencing data from more individuals are needed. The genomic data obtained provide additional reliable references for further comparative studies in hawkmoths (Sphingidae).

PMID:37550607 | DOI:10.1186/s12864-023-09506-y

Nat Nanotechnol. 2023 Aug 7. doi: 10.1038/s41565-023-01483-3. Online ahead of print.

ABSTRACT

Three-dimensional cell and organoid cultures rely on the mechanical support of viscoelastic matrices. However, commonly used matrix materials lack control over key cell-instructive properties. Here we report on fully synthetic hydrogels based on DNA libraries that self-assemble with ultrahigh-molecular-weight polymers, forming a dynamic DNA-crosslinked matrix (DyNAtrix). DyNAtrix enables computationally predictable and systematic control over its viscoelasticity, thermodynamic and kinetic parameters by changing DNA sequence information. Adjustable heat activation allows homogeneous embedding of mammalian cells. Intriguingly, stress-relaxation times can be tuned over four orders of magnitude, recapitulating mechanical characteristics of living tissues. DyNAtrix is self-healing, printable, exhibits high stability, cyto- and haemocompatibility, and controllable degradation. DyNAtrix-based cultures of human mesenchymal stromal cells, pluripotent stem cells, canine kidney cysts and human trophoblast organoids show high viability, proliferation and morphogenesis. DyNAtrix thus represents a programmable and versatile precision matrix for advanced approaches to biomechanics, biophysics and tissue engineering.

PMID:37550574 | DOI:10.1038/s41565-023-01483-3

Nat Microbiol. 2023 Aug 7. doi: 10.1038/s41564-023-01433-8. Online ahead of print.

ABSTRACT

Herpesviruses assemble large enveloped particles that are difficult to characterize structurally due to their size, fragility and complex multilayered proteome with partially amorphous nature. Here we used crosslinking mass spectrometry and quantitative proteomics to derive a spatially resolved interactome map of intact human cytomegalovirus virions. This enabled the de novo allocation of 32 viral proteins into four spatially resolved virion layers, each organized by a dominant viral scaffold protein. The viral protein UL32 engages with all layers in an N-to-C-terminal radial orientation, bridging nucleocapsid to viral envelope. We observed the layer-specific incorporation of 82 host proteins, of which 39 are selectively recruited. We uncovered how UL32, by recruitment of PP-1 phosphatase, antagonizes binding to 14-3-3 proteins. This mechanism assures effective viral biogenesis, suggesting a perturbing role of UL32-14-3-3 interaction. Finally, we integrated these data into a coarse-grained model to provide global insights into the native configuration of virus and host protein interactions inside herpesvirions.

PMID:37550507 | DOI:10.1038/s41564-023-01433-8

Nat Microbiol. 2023 Aug 7. doi: 10.1038/s41564-023-01448-1. Online ahead of print.

ABSTRACT

The fungal genus Armillaria contains necrotrophic pathogens and some of the largest terrestrial organisms that cause tremendous losses in diverse ecosystems, yet how they evolved pathogenicity in a clade of dominantly non-pathogenic wood degraders remains elusive. Here we show that Armillaria species, in addition to gene duplications and de novo gene origins, acquired at least 1,025 genes via 124 horizontal gene transfer events, primarily from Ascomycota. Horizontal gene transfer might have affected plant biomass degrading and virulence abilities of Armillaria, and provides an explanation for their unusual, soft rot-like wood decay strategy. Combined multi-species expression data revealed extensive regulation of horizontally acquired and wood-decay related genes, putative virulence factors and two novel conserved pathogenicity-induced small secreted proteins, which induced necrosis in planta. Overall, this study details how evolution knitted together horizontally and vertically inherited genes in complex adaptive traits of plant biomass degradation and pathogenicity in important fungal pathogens.

PMID:37550506 | DOI:10.1038/s41564-023-01448-1

Nat Med. 2023 Aug 7. doi: 10.1038/s41591-023-02474-6. Online ahead of print.

ABSTRACT

Treatment of circulatory shock in critically ill patients requires management of blood pressure using invasive monitoring, but uncertainty remains as to optimal individual blood pressure targets. Critical closing pressure, which refers to the arterial pressure when blood flow stops, can provide a fundamental measure of vascular tone in response to disease and therapy, but it has not previously been possible to measure this parameter routinely in clinical care. Here we describe a method to continuously measure critical closing pressure in the systemic circulation using readily available blood pressure monitors and then show that tissue perfusion pressure (TPP), defined as the difference between mean arterial pressure and critical closing pressure, provides unique information compared to other hemodynamic parameters. Using analyses of 5,988 admissions to a modern cardiac intensive care unit, and externally validated with 864 admissions to another institution, we show that TPP can predict the risk of mortality, length of hospital stay and peak blood lactate levels. These results indicate that TPP may provide an additional target for blood pressure optimization in patients with circulatory shock.

PMID:37550417 | DOI:10.1038/s41591-023-02474-6

Nat Plants. 2023 Aug 7. doi: 10.1038/s41477-023-01504-y. Online ahead of print.

NO ABSTRACT

PMID:37550373 | DOI:10.1038/s41477-023-01504-y

Sci Rep. 2023 Aug 7;13(1):12787. doi: 10.1038/s41598-023-39418-6.

ABSTRACT

We present an artificial neural network architecture, termed STENCIL-NET, for equation-free forecasting of spatiotemporal dynamics from data. STENCIL-NET works by learning a discrete propagator that is able to reproduce the spatiotemporal dynamics of the training data. This data-driven propagator can then be used to forecast or extrapolate dynamics without needing to know a governing equation. STENCIL-NET does not learn a governing equation, nor an approximation to the data themselves. It instead learns a discrete propagator that reproduces the data. It therefore generalizes well to different dynamics and different grid resolutions. By analogy with classic numerical methods, we show that the discrete forecasting operators learned by STENCIL-NET are numerically stable and accurate for data represented on regular Cartesian grids. A once-trained STENCIL-NET model can be used for equation-free forecasting on larger spatial domains and for longer times than it was trained for, as an autonomous predictor of chaotic dynamics, as a coarse-graining method, and as a data-adaptive de-noising method, as we illustrate in numerical experiments. In all tests, STENCIL-NET generalizes better and is computationally more efficient, both in training and inference, than neural network architectures based on local (CNN) or global (FNO) nonlinear convolutions.

PMID:37550328 | DOI:10.1038/s41598-023-39418-6