Biology subjects

Edgar, R. C.

Publications and source records attributed to Edgar, R. C..

At least 19 recordsLinked to original sources

bioRxiv2026

Rich structure alphabets enable highest accuracy protein search

Protein structure databases have grown from thousands of experimentally determined structures to hundreds of millions of AI-predicted models, creating an urgent need for search methods that combine high accuracy with practical scalability. Here, I present the third generation of Reseek, a protein structure search algorithm achieving the highest overall accuracy (median rank 1) according to diverse metrics among tested methods including DALI, Foldseek and TM-align. Improved accuracy is obtained by parallel sequence alignment of many discrete alphabets capturing primary, secondary and tertiary features, giving a combined space of[~] 1022 possible states. Separate statistical models are optimized for family, superfamily and fold discrimination, respectively, revealing distinct combinations of features that characterize each level. Hundreds of query structures can be searched a against a multimillion-structure database on a server computer in minutes, making large-scale structure search at state-of-the-art accuracy practical on commodity hardware.

bioinformatics↗

bioRxiv2025

Improved gene tree inference from removing alignment errors both from focal genes and when training substitution models

Multiple Sequence Alignment (MSA) is a key step in phylogenetic analysis, and is prone to error. Unfortunately, algorithms that remove likely alignment errors from MSAs sometimes also remove informative residues, making phylogenetic tree inference worse. Here we present a novel MSA cleaning algorithm based on consensus between MSAs using a range of Hidden Markov Models and guide trees, named CLOAK (CLeaning On the basis of Alignment C(K)onsensus). CLOAK is a gentle filter, with a low false positive rate for removal from MSAs according to the BALiBASE benchmarks, while still removing a significant fraction of likely alignment errors. Gentle vs. stringent MSA filtering methods are appropriate for different tasks. We assess methods based on their ability to bring the gene trees of single copy orthologs closer to the accepted species tree. Amino acid substitution models trained on filtered MSAs improve gene tree inference, with stricter filtering methods providing the biggest model improvements. In contrast, it is gentler filtering of single gene MSAs that provides additional improvements to gene tree inference, with CLOAK performing best.

evolutionary biology↗

bioRxiv2025

Protein structure alignment significance is often exaggerated

Machine learning has generated millions of high-quality predicted protein structures, creating a need for computationally efficient structure search algorithms and robust estimates of statistical significance at this scale. We show that unrelated proteins have a universal tendency towards convergent evolution of secondary and tertiary motifs, causing an excess of high-scoring false positive alignments. We investigate popular structure search and alignment algorithms, finding that previous methods routinely overestimate significance by up to six orders of magnitude. To address these issues, and to accommodate recent innovations in search algorithm design, we describe a novel method for estimating statistical significance. We show that its E-values are accurate, scale successfully with database size, and are robust against the (generally unknown) diversity of folds in the database. We implement our approach in an online structure search service based on Reseek at https://reseek.online.

bioinformatics↗

bioRxiv2024

Muscle-3D: scalable multiple protein structure alignment

Protein multiple alignment is an essential step in many bioinformatics analysis such as phylogenetic tree estimation, HMM construction and critical residue identification. Structure is conserved between distantly-related proteins where amino acid similarity is weak or undetectable, suggesting that structure-informed sequence alignments might offer advantages over alignments constructed from amino acid sequences alone. The advent of the AI folding era has unleashed millions of high-quality predicted structures, motivating the development and assessment of scalable multiple structure alignment (MStA) methods. Here, we describe Muscle-3D, a new MStA algorithm combining a rich sequence representation of structure context, the Reseek "mega-alphabet", with state-of-the art alignment techniques from Muscle5 including a posterior decoding pair-HMM, consistency transformation, iterative refinement and ensemble construction. We show that Muscle-3D readily scales to thousands of structures. Comparative validation on several benchmark datasets using different quality metrics shows Muscle-3D to be among the higher-scoring methods, but we find that algorithm rankings from different metrics disagree despite low P-values according to the Wilcoxon rank-sum test. We suggest that these conflicts arise from the inherently fuzzy nature of structural alignment, and argue that a universal standard of MStA accuracy is not possible in principle. We describe contact map profiles for visualizing variation in inter-residue distances, and introduce a novel measure of local conformation similarity, LDDT-muw. Muscle-3D software is available at https://github.com/rcedgar/muscle.

bioinformatics↗

bioRxiv2024

Logan: Planetary-Scale Genome Assembly Surveys Life's Diversity

The breadth of lifes diversity is unfathomable, but public nucleic acid sequencing data offers a window into the dispersion and evolution of genetic diversity across Earth. However the rapid growth and accumulation of sequence data have outpaced efficient analysis capabilities. The largest collection of freely available sequencing data is the Sequence Read Archive (SRA), comprising 27.3 million datasets or 5 x 1016 basepairs. To realize the potential of the SRA, we constructed Logan, a massive sequence assembly transforming short reads into long contigs and compressing the data over 100-fold, enabling highly efficient petabase-scale analysis. We created Logan-Search, a k-mer index of Logan for free planetary-scale sequence search, returning matches in minutes. We used Logan contigs to identify >200 million plastic-degrading enzyme homologs, and validate novel enzymes with catalytic activities exceeding current reference standards. Further, we vastly expand the known diversity of proteins (30-fold over UniRef50), plasmids (22-fold over PLSDB), P4 satellites (4.5-fold), and the recently described Obelisk RNA elements (3.7-fold). Logan also enables ecological and biomedical data mining, such as global tracking of antimicrobial resistance genes and the characterization of viral reactivation across millions of human BioSamples. By transforming the SRA, Logan democratizes access to the worlds public genetic data and opens frontiers in biotechnology, molecular ecology, and global health.

bioinformatics↗

bioRxiv2024

Sequence alignment using large protein structure alphabets doubles sensitivity to remote homologs

Recent breakthroughs in protein fold prediction from amino acid sequences have unleashed a deluge of new structures, raising new opportunities for expanding insights into the universe of proteins and pursuing practical applications in bio-engineering and therapeutics while also presenting new challenges to protein search and analysis algorithms. Here, I describe Reseek, a protein alignment algorithm which improves sensitivity in protein homolog detection compared to state-of-the-art methods including DALI, TM-align and Foldseek, with improved speed over Foldseek, the fastest previous method. Reseek is based on alignment of sequences where each residue in the protein backbone is represented by a letter in a novel "mega-alphabet" of 85,899,345,920 ([~] 1011) distinct states. Code is available at https://github.com/rcedgar/reseek.

bioinformatics↗

bioRxiv2024

Viroid-like colonists of human microbiomes

Here, we describe the "Obelisks," a previously unrecognised class of viroid-like elements that we first identified in human gut metatranscriptomic data. "Obelisks" share several properties: (i) apparently circular RNA [~]1kb genome assemblies, (ii) predicted rod-like secondary structures encompassing the entire genome, and (iii) open reading frames coding for a novel protein superfamily, which we call the "Oblins". We find that Obelisks form their own distinct phylogenetic group with no detectable sequence or structural similarity to known biological agents. Further, Obelisks are prevalent in tested human microbiome metatranscriptomes with representatives detected in [~]7% of analysed stool metatranscriptomes (29/440) and in [~]50% of analysed oral metatranscriptomes (17/32). Obelisk compositions appear to differ between the anatomic sites and are capable of persisting in individuals, with continued presence over >300 days observed in one case. Large scale searches identified 29,959 Obelisks (clustered at 90% nucleotide identity), with examples from all seven continents and in diverse ecological niches. From this search, a subset of Obelisks are identified to code for Obelisk-specific variants of the hammerhead type-III self-cleaving ribozyme. Lastly, we identified one case of a bacterial species (Streptococcus sanguinis) in which a subset of defined laboratory strains harboured a specific Obelisk RNA population. As such, Obelisks comprise a class of diverse RNAs that have colonised, and gone unnoticed in, human, and global microbiomes.

microbiology↗

bioRxiv2023

On-target, dual aminopeptidase inhibition provides cross-species antimalarial activity.

To combat the global burden of malaria, development of new drugs to replace or complement current therapies are urgently required. As drug resistance to existing treatments and clinical failures continue to rise, compounds targeting multiple life cycle stages and species need to be developed as a high priority. Here we show that the compound MMV1557817 is a nanomolar inhibitor of both Plasmodium falciparum and Plasmodium vivax aminopeptidases M1 and M17, leading to inhibition of end stage haemoglobin digestion in asexual parasites. Multi-stage analysis confirmed that MMV1557817 can also kill sexual stage P. falciparum, while cross-resistance studies confirmed the compound targets a mechanism of action distinct to current drug resistance mechanisms. Analysis of cross reactivity to homologous human enzymes shows the compound exhibits a high level of selectivity, whilst safety as well as druggability was confirmed in the murine model P. berghei. MMV1557817-resistant P. falciparum parasites displayed only low-level resistance (<3-fold) and exhibited a slow growth rate that was quickly outcompeted by wild type parasites. MMV1557817-resistant parasites digest significantly more haemoglobin and possess a mutation in PfA-M17 that induces partial destabilisation of the PfA-M17 homohexamer, resulting in high-level resistance to specific PfA-M17 inhibition, but enhanced sensitivity to specific PfA-M1 inhibition, and importantly, these parasites were highly sensitive to artemisinin. Overall, these results confirm MMV1557817 as a potential lead compound for further drug development and highlight the potential of dual inhibition of M1 and M17 as an effective multi-species drug targeting strategy.

microbiology↗

bioRxiv2022

Extant hybrids of RNA viruses and viroid-like elements

Earths life may have originated as self-replicating RNA. Some of the simplest current RNA replicators are RNA viruses, defined by linear RNA genomes encoding an RNA-dependent RNA polymerase (RdRP), and subviral agents with single-stranded, circular RNA genomes, such as viroids encoding paired self-cleaving ribozymes. Amongst a massive expansion of candidate viroid and viroid-like elements, we report that fungal pathogens, ambiviruses, are viroid-like elements which undergo rolling circle replication and encode their own viral RdRP, thus they are a distinct hybrid infectious agent. These findings point to a deep evolutionary history between modern RNA viruses and sub-viral elements and offer new perspectives on the evolution of primordial infectious agents, and RNA life. One-Sentence SummaryNovel infectious agents resembling self-cleaving viroid-like RNAs whilst encoding a viral RNA-dependent RNA polymerase.

microbiology↗

bioRxiv2022

Known phyla dominate the RNA virome in Tara Oceans data, not new phyla as claimed

The Tara Oceans project claimed to identify five new RNA virus phyla, two of which "are dominant in the oceans". However, their own assignments classify 28,353 of their putative RdRp-containing contigs to known phyla but only 886 (2.8%) to the five claimed new phyla combined. I mapped their reads to their contigs, finding that known phyla also account for a large majority (93.8%) of reads. I show that 510 of their putative viral contigs contain cellular proteins and further that predicted polymerase structures for their claimed phyla have incomplete and malformed palm domains, contradicting viral polymerase function and calling into question whether their putative polymerase sequences are derived from viruses.

microbiology↗

bioRxiv2021

MUSCLE v5 enables improved estimates of phylogenetic tree confidence by ensemble bootstrapping

Multiple sequence alignments (MSAs) are widely used to infer evolutionary relationships, enabling inferences of structure, function, and phylogeny. Standard practice is to construct one MSA by some preferred method and use it in further analysis; however, undetected MSA bias can be problematic. I describe Muscle5, a novel algorithm which constructs an ensemble of high-accuracy MSAs with diverse biases by perturbing a hidden Markov model and permuting its guide tree. Confidence in an inference is assessed as the fraction of the ensemble which supports it. Applied to phylogenetic tree estimation, I show that ensembles can confidently resolve topologies with low bootstrap according to standard methods, and conversely that some topologies with high bootstraps are incorrect. Applied to the phylogeny of RNA viruses, ensemble analysis shows that recently adopted taxonomic phyla are probably polyphyletic. Ensemble analysis can improve confidence assessment in any inference from an MSA.

bioinformatics↗

bioRxiv2021

Ribovirus classification by a polymerase barcode sequence

RNA viruses encoding a polymerase gene (riboviruses) dominate the known eukaryotic virome. Next-generation sequencing is revealing a wealth of new riboviruses with uncharacterised phenotypes, precluding classification by traditional taxonomic methods. These are often classified on the basis of polymerase sequence identity, but standardised methods to support this approach are currently lacking. To address this need, we describe the polymerase palmprint, a well-defined segment of the palm sub-domain delineated by well-conserved catalytic motifs. We present a novel algorithm, Palmscan, which identifies palmprints in nucleotide and amino acid sequences. We describe PALMdb, a reference database of palmprints derived from public sequence databases. Palmscan source code and PALMdb data are deposited at https://github.com/rcedgar/palmscan and https://github.com/rcedgar/palmdb, respectively.

bioinformatics↗

bioRxiv2020

Syncmers are more sensitive than minimizers for selecting conserved k-mers in biological sequences

Minimizers are widely used to select subsets of fixed-length substrings (k-mers) from biological sequences in applications ranging from read mapping to taxonomy prediction and indexing of large datasets. The minimizer of a string of w consecutive k-mers is the k-mer with smallest value according to an ordering of all k-mers. Syncmers are a family of alternative methods which select k-mers by inspecting the position of the smallest-valued substring of length s<k within the k-mer. For example, a closed syncmer selected if its smallest s-mer is at the start of end of the k-mer. At least one closed syncmer must be found in every window of length (k - s) k-mers. Unlike a minimizer, a syncmer is identified by its sequence alone, and is therefore synchronized in the following sense: if a given k-mer is selected from one sequence, it will also be selected from any other sequence. Also, minimizers can be deleted by mutations in flanking sequence, which cannot happen with syncmers. Experiments on minimizers with parameters used in the minimap2 read mapper and Kraken taxonomy prediction algorithm respectively show that syncmers can simultaneously achieve both lower density and higher conservation compared to minimizers.

bioinformatics↗

bioRxiv2020

Petabase-scale sequence alignment catalyses viral discovery

Public databases contain a planetary collection of nucleic acid sequences, but their systematic exploration has been inhibited by a lack of efficient methods for searching this corpus, now exceeding multiple petabases and growing exponentially [1, 2]. We developed a cloud computing infrastructure, Serratus, to enable ultra-high throughput sequence alignment at the petabase scale. We searched 5.7 million biologically diverse samples (10.2 petabases) for the hallmark gene RNA dependent RNA polymerase, identifying well over 105 novel RNA viruses and thereby expanding the number of known species by roughly an order of magnitude. We characterised novel viruses related to coronaviruses and to hepatitis{delta} virus, respectively and explored their environmental reservoirs. To catalyse a new era of viral discovery, we established a free and comprehensive database of these data and tools. Expanding the known sequence diversity of viruses can reveal the evolutionary origins of emerging pathogens and improve pathogen surveillance for the anticipation and mitigation of future pandemics.

bioinformatics↗

bioRxiv2020

Capturing and Recreating Diverse Antibody Repertoires as Multivalent Recombinant Polyclonal Antibody Drugs

Plasma-derived polyclonal antibodies are polyvalent drugs used for many important clinical indications that require modulation of multiple drug targets simultaneously, including emerging infectious disease and transplantation. However, plasma-derived drugs suffer many problems, including low potency, impurities, constraints on supply, and batch-to-batch variation. In this study, we demonstrated proofs-of-concept for a technology that uses microfluidics and molecular genomics to capture diverse mammalian antibody repertoires as multivalent recombinant drugs. These "recombinant hyperimmune" drugs comprised thousands to tens of thousands of antibodies and were derived from convalescent human donors, or vaccinated human donors or immunized mice. Here we used our technology to build a highly potent recombinant hyperimmune for Severe Acute Respiratory Syndrome Coronavirus-2 (SARS CoV-2) in less than three months. We also validated a recombinant hyperimmune for Zika virus disease that abrogates antibody-dependent enhancement (ADE) through Fc engineering. For patients with primary immune deficiency (PID), we built high potency polyvalent recombinant hyperimmunes against pathogens that commonly cause serious lung infections. Finally, to address the limitations of rabbit-derived anti-thymocyte globulin (ATG), we generated a recombinant human version and demonstrated in vivo function against graft-versus-host disease (GVHD). Recombinant hyperimmunes are a novel class of drugs that could be used to target a wide variety of other clinical applications, including cancer and autoimmunity.

immunology↗

bioRxiv2020

URMAP, an ultra fast read mapper

Mapping of reads to reference sequences is an essential step in a wide range of biological studies. The large size of datasets generated with next-generation sequencing technologies motivates the development of fast mapping software. Here, I describe URMAP, a new read mapping algorithm. URMAP is an order of magnitude faster than BWA and Bowtie2 with comparable accuracy on a benchmark test using simulated paired 150nt reads of a well-studied human genome. Software is freely available at https://drive5.com/urmap.

bioinformatics↗

bioRxiv2018

Alpha diversity metrics for noisy OTUs

Next-generation sequencing (NGS) of marker genes such as 16S ribosomal RNA is widely used to survey microbial communities. The in-sample (alpha) diversity of Operational Taxonomic Units (OTUs) is often summarized by metrics such as richness or entropy which are calculated from observed abundances, or by estimators such as Chao1 which extrapolate to unobserved OTUs. Most such measures are adopted from traditional biodiversity studies, where observational error can often be neglected. However, errors introduced by next-generation amplicon sequencing tend to induce spurious OTUs and spurious counts in OTU tables, both of which are especially prevalent at low abundances. In consequence, traditional metrics may be grossly inaccurate if they are naively applied to NGS OTU tables. In this work, we describe two novel alpha diversity estimators which are calculated from OTU abundances above a specified threshold. The singleton-free estimator (SFE) is a non-parametric estimator which is derived from a similar approach to Chao1 but extrapolates using doublet and triplet abundances rather than singletons and doublets. The octave estimator (OE) fits a log-normal distribution to non-singleton bars of an octave plot. We show that these estimators are effective under suitable conditions, but these conditions rarely apply in practice. We conclude that extrapolating to unobserved OTUs remains an open problem which is unlikely to be solved in the near future.

bioinformatics↗

bioRxiv2018

Octave plots for visualizing diversity of microbial OTUs

Next-generation sequencing of marker genes such as 16S ribosomal RNA is widely used to survey microbial communities. The abundance distribution (AD) of Operational Taxonomic Units (OTUs) in a sample is typically summarized by alpha diversity metrics, e.g. richness and entropy, discarding information about the AD shape. In this work, we describe octave plots, histograms which visualize the shape of microbial ADs by binning on a logarithmic scale with base 2. Optionally, histogram bars are colored to indicate possible spurious OTUs due to sequence error and cross-talk. Octave plots enable assessment of (a) the shape and completeness of the distribution, (b) the effects of noise on measured diversity, (c) whether low-abundance OTUs should be discarded, (d) whether alpha diversity metrics and estimators are reliable, and (e) the additional sampling effort (i.e., read depth) required to obtain a complete census of the community. The utility of octave plots is illustrated in a re-analysis of a prostate cancer study showing that the reported core microbiome is most likely an artifact of experimental error.

bioinformatics↗