WORMHOLE: Novel Least Diverged Ortholog Prediction through Machine Learning

SonicParanoid2: fast, accurate, and comprehensive orthology inference with machine learning and language models

Accurate inference of orthologous genes constitutes a prerequisite for comparative and evolutionary genomics. SonicParanoid is one of the fastest tools for orthology inference; however, its scalability and accuracy have been hampered by time-consuming all-versus-all alignments and the existence of proteins with complex domain architectures. Here, we present a substantial update of SonicParanoid, where a gradient boosting predictor halves the execution time and a language model doubles the recall. Application to empirical large-scale and standardized benchmark datasets shows that SonicParanoid2 is much faster than comparable methods and also the most accurate. SonicParanoid2 is available at https://gitlab.com/salvo981/sonicparanoid2 and https://zenodo.org/doi/10.5281/zenodo.11371108 .

SERBP1 interacts with PARP1 and is present in PARylation-dependent protein complexes regulating splicing, cell division, and ribosome biogenesis

RNA binding proteins (RBPs) containing intrinsically disordered regions (IDRs) are present in diverse molecular complexes where they function as dynamic regulators. Their characteristics promote liquid-liquid phase separation (LLPS) and the formation of membraneless organelles such as stress granules and nucleoli. IDR-RBPs are particularly relevant in the nervous system and their dysfunction is associated with neurodegenerative diseases and brain tumor development. SERBP1 is a unique member of this group, being mostly disordered and lacking canonical RNA-binding domains. Using a proteomics approach followed by functional analysis, we defined SERBP1's interactome. We uncovered novel SERBP1 roles in splicing, cell division, and ribosomal biogenesis and showed its participation in pathological stress granules and Tau aggregates in Alzheimer's disease brains. SERBP1 preferentially interacts with other G-quadruplex (G4) binders, implicated in different stages of gene expression, suggesting that G4 binding is a critical component of SERBP1 function in different settings. Similarly, we identified important associations between SERBP1 and PARP1/polyADP-ribosylation (PARylation). SERBP1 interacts with PARP1 and its associated factors and influences PARylation. Moreover, protein complexes in which SERBP1 participates contain mostly PARylated proteins and PAR binders. Based on these results, we propose a feedback regulatory model in which SERBP1 influences PARP1 function and PARylation, while PARylation modulates SERBP1 functions and participation in regulatory complexes.

Inferring Interaction Networks from Transcriptomic Data: Methods and Applications

Transcriptomic data is a treasure trove in modern molecular biology, as it offers a comprehensive viewpoint into the intricate nuances of gene expression dynamics underlying biological systems. This genetic information must be utilized to infer biomolecular interaction networks that can provide insights into the complex regulatory mechanisms underpinning the dynamic cellular processes. Gene regulatory networks and protein-protein interaction networks are two major classes of such networks. This chapter thoroughly investigates the wide range of methodologies used for distilling insightful revelations from transcriptomic data that include association-based methods (based on correlation among expression vectors), probabilistic models (using Bayesian and Gaussian models), and interologous methods. We reviewed different approaches for evaluating the significance of interactions based on the network topology and biological functions of the interacting molecules and discuss various strategies for the identification of functional modules. The chapter concludes with highlighting network-based techniques of prioritizing key genes, outlining the centrality-based, diffusion- based, and subgraph-based methods. The chapter provides a meticulous framework for investigating transcriptomic data to uncover assembly of complex molecular networks for their adaptable analyses across a broad spectrum of biological domains.

Joint representation of molecular networks from multiple species improves gene classification

Network-based machine learning (ML) has the potential for predicting novel genes associated with nearly any health and disease context. However, this approach often uses network information from only the single species under consideration even though networks for most species are noisy and incomplete. While some recent methods have begun addressing this shortcoming by using networks from more than one species, they lack one or more key desirable properties: handling networks from more than two species simultaneously, incorporating many-to-many orthology information, or generating a network representation that is reusable across different types of and newly-defined prediction tasks. Here, we present GenePlexusZoo, a framework that casts molecular networks from multiple species into a single reusable feature space for network-based ML. We demonstrate that this multi-species network representation improves both gene classification within a single species and knowledge-transfer across species, even in cases where the inter-species correspondence is undetectable based on shared orthologous genes. Thus, GenePlexusZoo enables effectively leveraging the high evolutionary molecular, functional, and phenotypic conservation across species to discover novel genes associated with diverse biological contexts.

On the benefits of the tryptophan metabolite 3-hydroxyanthranilic acid in Caenorhabditis elegans and mouse aging

Tryptophan metabolism through the kynurenine pathway influences molecular processes critical to healthy aging including immune signaling, redox homeostasis, and energy production. Aberrant kynurenine metabolism occurs during normal aging and is implicated in many age-associated pathologies including chronic inflammation, atherosclerosis, neurodegeneration, and cancer. We and others previously identified three kynurenine pathway genes-tdo-2, kynu-1, and acsd-1-for which decreasing expression extends lifespan in invertebrates. Here we report that knockdown of haao-1, a fourth gene encoding the enzyme 3-hydroxyanthranilic acid (3HAA) dioxygenase (HAAO), extends lifespan by ~30% and delays age-associated health decline in Caenorhabditis elegans. Lifespan extension is mediated by increased physiological levels of the HAAO substrate 3HAA. 3HAA increases oxidative stress resistance and activates the Nrf2/SKN-1 oxidative stress response. In pilot studies, female Haao knockout mice or aging wild type male mice fed 3HAA supplemented diet were also long-lived. HAAO and 3HAA represent potential therapeutic targets for aging and age-associated disease.

Unfolded protein response IRE1/XBP1 signaling is required for healthy mammalian brain aging

Aging is a major risk factor to develop neurodegenerative diseases and is associated with decreased buffering capacity of the proteostasis network. We investigated the significance of the unfolded protein response (UPR), a major signaling pathway activated to cope with endoplasmic reticulum (ER) stress, in the functional deterioration of the mammalian brain during aging. We report that genetic disruption of the ER stress sensor IRE1 accelerated age-related cognitive decline. In mouse models, overexpressing an active form of the UPR transcription factor XBP1 restored synaptic and cognitive function, in addition to reducing cell senescence. Proteomic profiling of hippocampal tissue showed that XBP1 expression significantly restore changes associated with aging, including factors involved in synaptic function and pathways linked to neurodegenerative diseases. The genes modified by XBP1 in the aged hippocampus where also altered. Collectively, our results demonstrate that strategies to manipulate the UPR in mammals may help sustain healthy brain aging.

A Brain Tumor Image Segmentation Method Based on Quantum Entanglement and Wormhole Behaved Particle Swarm Optimization

Purpose

Although classical techniques for image segmentation may work well for some images, they may perform poorly or not work at all for others. It often depends on the properties of the particular image segmentation task under study. The reliable segmentation of brain tumors in medical images represents a particularly challenging and essential task. For example, some brain tumors may exhibit complex so-called “bottle-neck” shapes which are essentially circles with long indistinct tapering tails, known as a “dual tail.” Such challenging conditions may not be readily segmented, particularly in the extended tail region or around the so-called “bottle-neck” area. In those cases, existing image segmentation techniques often fail to work well.

Methods

Existing research on image segmentation using wormhole and entangle theory is first analyzed. Next, a random positioning search method that uses a quantum-behaved particle swarm optimization (QPSO) approach is improved by using a hyperbolic wormhole path measure for seeding and linking particles. Finally, our novel quantum and wormhole-behaved particle swarm optimization (QWPSO) is proposed.

Results

Experimental results show that our QWPSO algorithm can better cluster complex “dual tail” regions into groupings with greater adaptability than conventional QPSO. Experimental work also improves operational efficiency and segmentation accuracy compared with current competing reference methods.

Conclusion

Our QWPSO method appears extremely promising for isolating smeared/indistinct regions of complex shape typical of medical image segmentation tasks. The technique is especially advantageous for segmentation in the so-called “bottle-neck” and “dual tail”-shaped regions appearing in brain tumor images.

The Quest for Orthologs orthology benchmark service in 2022

The Orthology Benchmark Service (https://orthology.benchmarkservice.org) is the gold standard for orthology inference evaluation, supported and maintained by the Quest for Orthologs consortium. It is an essential resource to compare existing and new methods of orthology inference (the bedrock for many comparative genomics and phylogenetic analysis) over a standard dataset and through common procedures. The Quest for Orthologs Consortium is dedicated to maintaining the resource up to date, through regular updates of the Reference Proteomes and increasingly accessible data through the OpenEBench platform. For this update, we have added a new benchmark based on curated orthology assertion from the Vertebrate Gene Nomenclature Committee, and provided an example meta-analysis of the public predictions present on the platform.

Ten Years of Collaborative Progress in the Quest for Orthologs

Accurate determination of the evolutionary relationships between genes is a foundational challenge in biology. Homology-evolutionary relatedness-is in many cases readily determined based on sequence similarity analysis. By contrast, whether or not two genes directly descended from a common ancestor by a speciation event (orthologs) or duplication event (paralogs) is more challenging, yet provides critical information on the history of a gene. Since 2009, this task has been the focus of the Quest for Orthologs (QFO) Consortium. The sixth QFO meeting took place in Okazaki, Japan in conjunction with the 67th National Institute for Basic Biology conference. Here, we report recent advances, applications, and oncoming challenges that were discussed during the conference. Steady progress has been made toward standardization and scalability of new and existing tools. A feature of the conference was the presentation of a panel of accessible tools for phylogenetic profiling and several developments to bring orthology beyond the gene unit-from domains to networks. This meeting brought into light several challenges to come: leveraging orthology computations to get the most of the incoming avalanche of genomic data, integrating orthology from domain to biological network levels, building better gene models, and adapting orthology approaches to the broad evolutionary and genomic diversity recognized in different forms of life and viruses.

Composition of Caenorhabditis elegans extracellular vesicles suggests roles in metabolism, immunity, and aging

The nematode Caenorhabditis elegans has been instrumental in the identification of evolutionarily conserved mechanisms of aging. C. elegans also has recently been found to have evolutionarily conserved extracellular vesicle (EV) signaling pathways. We have been developing tools that allow for the detailed study of EV biology in C. elegans. Here we apply our recently published method for high specificity purification of EVs from C. elegans to carry out target-independent proteomic and RNA analysis of nematode EVs. We identify diverse coding and non-coding RNA and protein cargo types commonly found in human EVs. The EV cargo spectrum is distinct from whole worms, suggesting that protein and RNA cargos are actively recruited to EVs. Gene ontology analysis revealed C. elegans EVs are enriched for extracellular-associated and signaling proteins, and network analysis indicates enrichment for metabolic, immune, and basement membrane associated proteins. Tissue enrichment and gene expression analysis suggests the secreted EV proteins are likely to be derived from intestine, muscle, and excretory tissue. An unbiased comparison of the EV proteins with a large database of C. elegans genome-wide microarray data showed significant overlap with gene sets that are associated with aging and immunity. Taken together our data suggest C. elegans could be a promising in vivo model for studying the genetics and physiology of EVs in a variety of contexts including aging, metabolism, and immune response.

Advances and Applications in the Quest for Orthologs

Gene families evolve by the processes of speciation (creating orthologs), gene duplication (paralogs), and horizontal gene transfer (xenologs), in addition to sequence divergence and gene loss. Orthologs in particular play an essential role in comparative genomics and phylogenomic analyses. With the continued sequencing of organisms across the tree of life, the data are available to reconstruct the unique evolutionary histories of tens of thousands of gene families. Accurate reconstruction of these histories, however, is a challenging computational problem, and the focus of the Quest for Orthologs Consortium. We review the recent advances and outstanding challenges in this field, as revealed at a symposium and meeting held at the University of Southern California in 2017. Key advances have been made both at the level of orthology algorithm development and with respect to coordination across the community of algorithm developers and orthology end-users. Applications spanned a broad range, including gene function prediction, phylostratigraphy, genome evolution, and phylogenomics. The meetings highlighted the increasing use of meta-analyses integrating results from multiple different algorithms, and discussed ongoing challenges in orthology inference as well as the next steps toward improvement and integration of orthology resources.

Computational tools for geroscience

The rapid progress of the past three decades has led the geroscience field near a point where human interventions in aging are plausible. Advances across scientific areas, such as high throughput "-omics" approaches, have led to an exponentially increasing quantity of data available for biogerontologists. To best translate the lifespan and healthspan extending interventions discovered by basic scientists into preventative medicine, it is imperative that the current data are comprehensively utilized to generate testable hypotheses about translational interventions. Building a translational pipeline for geroscience will require both systematic efforts to identify interventions that extend healthspan across taxa and diagnostics that can identify patients who may benefit from interventions prior to the onset of an age-related morbidity. Databases and computational tools that organize and analyze both the wealth of information available on basic biogerontology research and clinical data on aging populations will be critical in developing such a pipeline. Here, we review the current landscape of databases and computational resources available for translational aging research. We discuss key platforms and tools available for aging research, with a focus on how each tool can be used in concert with hypothesis driven experiments to move closer to human interventions in aging.

Assigning confidence scores to homoeologs using fuzzy logic

In polyploid genomes, homoeologs are a specific subtype of homologs, and can be thought of as orthologs between subgenomes. In Orthologous MAtrix, we infer homoeologs in three polyploid plant species: upland cotton (Gossypium hirsutum), rapeseed (Brassica napus), and bread wheat (Triticum aestivum). While we can typically recognize the features of a "good" homoeolog prediction (a consistent evolutionary distance, high synteny, and a one-to-one relationship), none of them is a hard-fast criterion. We devised a novel fuzzy logic-based method to assign confidence scores to each pair of predicted homoeologs. We inferred homoeolog pairs and used the new and improved method to assign confidence scores, which ranged from 0 to 100. Most confidence scores were between 70 and 100, but the distribution varied between genomes. The new confidence scores show an improvement over our previous method and were manually evaluated using a subset from various confidence ranges.

Inferring Orthology and Paralogy

The distinction between orthologs and paralogs, genes that started diverging by speciation versus duplication, is relevant in a wide range of contexts, most notably phylogenetic tree inference and protein function annotation. In this chapter, we provide an overview of the methods used to infer orthology and paralogy. We survey both graph-based approaches (and their various grouping strategies) and tree-based approaches, which solve the more general problem of gene/species tree reconciliation. We discuss conceptual differences among the various orthology inference methods and databases and examine the difficult issue of verifying and benchmarking orthology predictions. Finally, we review typical applications of orthologous genes, groups, and reconciled trees and conclude with thoughts on future methodological developments.

OrthoList 2: A New Comparative Genomic Analysis of Human and Caenorhabditis elegans Genes

OrthoList, a compendium of Caenorhabditis elegans genes with human orthologs compiled in 2011 by a meta-analysis of four orthology-prediction methods, has been a popular tool for identifying conserved genes for research into biological and disease mechanisms. However, the efficacy of orthology prediction depends on the accuracy of gene-model predictions, an ongoing process, and orthology-prediction algorithms have also been updated over time. Here we present OrthoList 2 (OL2), a new comparative genomic analysis between C. elegans and humans, and the first assessment of how changes over time affect the landscape of predicted orthologs between two species. Although we find that updates to the orthology-prediction methods significantly changed the landscape of C. elegans-human orthologs predicted by individual programs and-unexpectedly-reduced agreement among them, we also show that our meta-analysis approach "buffered" against changes in gene content. We show that adding results from more programs did not lead to many additions to the list and discuss reasons to avoid assigning "scores" based on support by individual orthology-prediction programs; the treatment of "legacy" genes no longer predicted by these programs; and the practical difficulties of updating due to encountering deprecated, changed, or retired gene identifiers. In addition, we consider what other criteria may support claims of orthology and alternative approaches to find potential orthologs that elude identification by these programs. Finally, we created a new web-based tool that allows for rapid searches of OL2 by gene identifiers, protein domains [InterPro and SMART (Simple Modular Architecture Research Tool], or human disease associations ([OMIM (Online Mendelian Inheritence in Man], and also includes available RNA-interference resources to facilitate potential translational cross-species studies.

Caenorhabditis elegans orthologs of human genes differentially expressed with age are enriched for determinants of longevity

We report a systematic RNAi longevity screen of 82 Caenorhabditis elegans genes selected based on orthology to human genes differentially expressed with age. We find substantial enrichment in genes for which knockdown increased lifespan. This enrichment is markedly higher than published genomewide longevity screens in C. elegans and similar to screens that preselected candidates based on longevity‐correlated metrics (e.g., stress resistance). Of the 50 genes that affected lifespan, 46 were previously unreported. The five genes with the greatest impact on lifespan (>20% extension) encode the enzyme kynureninase (kynu‐1), a neuronal leucine‐rich repeat protein (iglr‐1), a tetraspanin (tsp‐3), a regulator of calcineurin (rcan‐1), and a voltage‐gated calcium channel subunit (unc‐36). Knockdown of each gene extended healthspan without impairing reproduction. kynu‐1(RNAi) alone delayed pathology in C. elegans models of Alzheimer's disease and Huntington's disease. Each gene displayed a distinct pattern of interaction with known aging pathways. In the context of published work, kynu‐1, tsp‐3, and rcan‐1 are of particular interest for immediate follow‐up. kynu‐1 is an understudied member of the kynurenine metabolic pathway with a mechanistically distinct impact on lifespan. Our data suggest that tsp‐3 is a novel modulator of hypoxic signaling and rcan‐1 is a context‐specific calcineurin regulator. Our results validate C. elegans as a comparative tool for prioritizing human candidate aging genes, confirm age‐associated gene expression data as valuable source of novel longevity determinants, and prioritize select genes for mechanistic follow‐up.