Biology subjects

Lucaci, A. G.

Publications and source records attributed to Lucaci, A. G..

10 recordsLinked to original sources

bioRxiv2026

HyphAeon: Attention on Evolution Across Deep Time Transforms Comparative Genomics

Detecting Darwinian natural selection is fundamental to evolutionary biology and functional genomics, yet standard methods based on phylogenetic models that estimate the ratio of non-synonymous to synonymous substitution rates (dN/dS) fail to scale with modern genomic volumes. Fitting continuous-time Markov substitution matrices across dense trees with hundreds of species requires extensive compute, forcing comparative genomics to rely on aggressive taxon subsampling or static whole-tree summaries that dilute transient adaptive bursts. Here we present HyphAeon, a lightweight (~1.91M parameter backbone, 2.46M across the full multi-task suite) phylogeny-informed foundation transformer trained to amortize the detection of episodic diversifying selection across 742-species mammalian coding alignments (17,186 genes, 9.77 x 10^6 codons). HyphAeon approaches the discriminative accuracy of numerical maximum-likelihood selection tests (MEME) across episodic burst regimes (ROC-AUC up to 0.942, mean 0.659; empirical Precision-Recall lift up to 25.8x, mean 7.7x; rank concordance up to rho = 0.983) while executing >1,000x faster per locus (averaging <1 ms per site across genome-wide scans) and >10,000x faster at proteome scale, generalizing outside its mammalian training distribution without retraining. Beyond accelerating classical tests, embedding molecular evolution into a differentiable geometric latent space enables analytical capabilities inaccessible to static dN/dS models: (1) targeted alignment artifact correction via counterfactual attribution; (2) macromolecular contact recovery and multi-site epistatic sectors (CESI); (3) directional phenotype-to-genotype attribution in lineage space (PARS); and (4) continuous temporal surveillance regression that tracks positive sweep velocities across longitudinal cohorts (evaluated across 12,167 timestamped genomes and benchmarked against external frequencies from >9.34 million genomes), rescuing adaptive substitutions obscured by post-fixation dilution. By bridging statistical phylogenetics with geometric representation learning, HyphAeon establishes comparative genomics as an interactive, high-throughput computational framework for evolutionary discovery.

bioinformatics↗

bioRxiv2026

Leveraging a hybrid cross-disciplinary training model to accelerate global bioinformatics capacity

Disparities in formal bioinformatics training exacerbate the global skills gap, impeding the democratized application of advanced genomic technologies. To bridge this divide, we introduce a scalable, hybrid training framework designed to rapidly accelerate regional bioinformatics capacity. We exemplify this approach through the Eastern European Bioinformatics and Genomics (EEBG) workshop series -- a cross-disciplinary initiative that pairs international faculty with local institutions to deliver modular, hands-on curricula. Functioning as a structured knowledge-transfer pipeline, the series has catalyzed a sustainable educational ecosystem, evidenced by the establishment of multiple independent summer schools across the region. The assessment of the 2025 EEBG workshop in Krakow, Poland, validates the models viability; participant metrics confirm high efficacy in skill acquisition (mean satisfaction: 4.4/5.0) and community building. Crucially, the hybrid delivery mode dismantled geographic barriers, serving as a vital mechanism for maintaining scientific continuity for researchers facing displacement and crisis. Synthesizing these outcomes, we define the core features of a replicable blueprint for scientific readiness in resource-constrained environments. We conclude by presenting a strategic roadmap -- organized around infrastructure standardization, governance sustainability, and geographical expansion -- for adapting this regional proof-of-concept into a global export-ready model, offering a critical path toward ensuring universal access to genomic innovation.

scientific communication and education↗

bioRxiv2023

Evolutionary analysis of mammalian Rem2, a member of the RGK (Rem, Rem2, Rad, and Gem/Kir) family of small GTPases, reveals the role of selection and epistasis in shaping protein functional constraints.

Rad And Gem-Like GTP-Binding Protein 2 (Rem2), is a member of the RGK family of Ras-like GTPases and has been identified in various mammalian species. Rem2 has been implicated in Huntingtons disease and Long QT Syndrome and is highly expressed in the brain and in endocrine cells. In this study, we examined the evolutionary history of Rem2 across mammals, focusing on the role of purifying selection and epistasis in shaping its sequence and structure. In our analysis of Rem2 sequences across 175 mammalian species, we found evidence for strong purifying selection in 70% of non-invariant codon sites of the protein, characteristic of essential proteins that play critical roles in biological processes and is consistent with Rem2s role in the regulation of neuronal development and function. We inferred epistatic effects in 49 pairs of coevolving codon sites in Rem2, some of which are predicted to have deleterious effects on human health. Additionally, we reconstructed the ancestral evolutionary history of mammalian Rem2 using protein structure prediction of extinct and extant sequences. This analysis revealed the dynamics of how substitutions that change the genetic distance of Rem2 can impact protein structure in variable regions while maintaining core functional mechanisms. By understanding the selective pressures, protein- and genetic-interactions that have shaped the sequence and structure of the Rem2 protein, we may gain a stronger understanding of its biological and functional constraints.

neuroscience↗

bioRxiv2023

Natural selection differences detected in key protein domains between non-pathogenic and pathogenic Feline Coronavirus phenotypes

Feline Coronaviruses (FCoVs) commonly cause mild enteric infections in felines worldwide (termed Feline Enteric Coronavirus [FECV]), with around 12% developing into deadly Feline Infectious Peritonitis (FIP; Feline Infectious Peritonitis Virus [FIPV]). Genomic differences between FECV and FIPV have been reported, yet the putative genotypic basis of the highly pathogenic phenotype remains unclear. Here, we used state-of-the-art molecular evolutionary genetic statistical techniques to identify and compare differences in natural selection pressure between FECV and FIPV sequences, as well as to identify FIPV and FECV specific signals of positive selection. We analyzed full length FCoV protein coding genes thought to contain mutations associated with FIPV (Spike, ORF3abc, and ORF7ab). We identified two sites exhibiting differences in natural selection pressure between FECV and FIPV: one within the S1/S2 furin cleavage site, and the other within the fusion domain of Spike. We also found 15 sites subject to positive selection associated with FIPV within Spike, 11 of which have not previously been suggested as possibly relevant to FIP development. These sites fall within Spike protein subdomains that participate in host cell receptor interaction, immune evasion, tropism shifts, host cellular entry, and viral escape. There were 14 sites (12 novel) within Spike under positive selection associated with the FECV phenotype, almost exclusively within the S1/S2 furin cleavage site and adjacent C domain, along with a signal of relaxed selection in FIPV relative to FECV, suggesting that furin cleavage functionality may not be needed for FIPV. Positive selection inferred in ORF7b was associated with the FECV phenotype, and included 24 positively selected sites, while ORF7b had signals of relaxed selection in FIPV. We found evidence of positive selection in ORF3c in FCoV wide analyses, but no specific association with the FIPV or FECV phenotype. We hypothesize that some combination of mutations in FECV may contribute to FIP development, and that is unlikely to be one singular "switch" mutational event. This work expands our understanding of the complexities of FIP development and provides insights into how evolutionary forces may alter pathogenesis in coronavirus genomes.

evolutionary biology↗

bioRxiv2022

Evolutionary shortcuts via multi-nucleotide substitutions and their impact on natural selection analyses.

Inference and interpretation of evolutionary processes, in particular of the types and targets of natural selection affecting coding sequences, are critically influenced by the assumptions built into statistical models and tests. If certain aspects of the substitution process (even when they are not of direct interest) are presumed absent or are modeled with too crude of a simplification, estimates of key model parameters can become biased, often systematically, and lead to poor statistical performance. Previous work established that failing to accommodate multi-nucleotide (or multi-hit, MH) substitutions strongly biases dN/dS-based inference towards false positive inferences of diversifying episodic selection, as does failing to model variation in the rate of synonymous substitution (SRV) among sites. Here we develop an integrated analytical framework and software tools to simultaneously incorporate these sources of evolutionary complexity into selection analyses. We found that both MH and SRV are ubiquitous in empirical alignments, and incorporating them has a strong effect on whether or not positive selection is detected, (1.4-fold reduction) and on the distributions of inferred evolutionary rates. With simulation studies, we show that this effect is not attributable to reduced statistical power caused by using a more complex model. After a detailed examination of 21 benchmark alignments and a new high-resolution analysis showing which parts of the alignment provide support for positive selection, we show that MH substitutions occurring along shorter branches in the tree explain a significant fraction of discrepant results in selection detection. Our results add to the growing body of literature which examines decadesold modeling assumptions (including MH) and finds them to be problematic for comparative genomic data analysis. Because multi-nucleotide substitutions have a significant impact on natural selection detection even at the level of an entire gene, we recommend that selection analyses of this type consider their inclusion as a matter of routine. To facilitate this procedure, we developed, implemented, and benchmarked a simple and well-performing model testing selection detection framework able to screen an alignment for positive selection with two biologically important confounding processes: site-to-site synonymous rate variation, and multi-nucleotide instantaneous substitutions.

bioinformatics↗

bioRxiv2022

Human HspB1, HspB3, HspB5 and HspB8: Shaping these Disease Factors during Vertebrate Evolution

Small heat shock proteins (sHSPs) emerged early in evolution and occur in all domains of life and nearly in all species, including humans. Mutations in four sHSPs (HspB1, HspB3, HspB5, HspB8) are associated with neuromuscular disorders. The aim of this study is to investigate the evolutionary forces shaping these sHSPs during vertebrate evolution. We performed comparative evolutionary analyses on a set of orthologous sHSP sequences, based on the ratio of non-synonymous: synonymous substitution rates for each codon. We found that these sHSPs had been historically exposed to different degrees of purifying selection, decreasing in this order: HspB8 > HspB1, HspB5 > HspB3. Within each sHSP, regions with different degrees of purifying selection can be discerned, resulting in characteristic selective pressure profiles. The conserved -crystallin domains were exposed to the most stringent purifying selection compared to the flanking regions, supporting a dimorphic pattern of evolution. Thus, during vertebrate evolution the different sequence partitions were exposed to different and measurable degrees of selective pressures. Among the disease-associated mutations, most are missense mutations primarily in HspB1 and to a minor extent in the other sHSPs. Our data provide an explanation for this disparate incidence. Contrary to the expectation, most missense mutations cause dominant disease phenotypes. Theoretical considerations support a connection between the historic exposure of these sHSP genes to a high degree of purifying selection and the unusual prevalence of genetic dominance of the associated disease phenotypes. Our study puts the genetics of inheritable sHSP-borne diseases into the context of vertebrate evolution.

evolutionary biology↗

bioRxiv2022

The evolution of BDNF is defined by strict purifying selection and prodomain spatial coevolution, but what does it mean for human brain disease?

The mammalian gene Brain-Derived Neurotrophic Factor (BDNF) is an essential mediator of brain assembly, development, and maturation which has been implicated in a variety of brain disorders such as neurodevelopmental disorders (e.g. autism spectrum disorder), neuropsychiatric disorders (e.g. depression, PTSD, schizophrenia), and neurodegenerative disorders (e.g. Parkinsons). Loss of BDNF during early development is embryonic lethal, and depletion of BDNF during adolescence or adulthood can result in disease-related neuropathology across a broad range of model organisms. In order to better understand the role of BDNF in disease, we seek to provide an evolutionary context to BDNFs role within the brain by elucidating the molecular and genetic comparative history of BDNF across species. We conduct sequence alignment and phylogenetic reconstruction of the BDNF gene across a diverse selection of over 160 mammalian species spanning [~]177 million years of evolution. Selective evolutionary change was examined via several independent computational models of codon evolution including FEL, MEME, and BGM. We report strict purifying selection in the main functional domain of BDNF (NGF domain, essentially comprising the mature BDNF protein). Specifically, we discover 6 sites in our homologous alignment which are under episodic selection in the early regulatory region of BDNF (i.e. the prodomain) and 23 pairs of coevolving sites that are a part of complex spatial relationships that are distributed across the entire BDNF gene. Thus, we propose that our discovery of both local and distal sites of co-evolution within the pro- and mature-domains of BDNF that likely reflect the evolutionary fine-tuning of BDNFs unique and complex regulatory capacities whilst also retaining its core yet diverse ontogenic functionality within the central nervous system. This discovery consequently supports the idea that the BDNF prodomain is more prone to change than the mature domain, however the fact that this region has also been subject to negative purifying selection also highlights genetic sensitivity and thus partially explains the prodomains disease relevance (e.g. Val66Met and other variants) to numerous neuropsychiatric disorders. HIGHLIGHTSO_LIWe extracted coding sequences for Brain-Derived Neurotrophic Factor (BDNF) from over 160 mammalian genomes that span approximately [~]177 million years of evolution. C_LIO_LIWe observe strict purifying selection in the main functional domain (NGF) of the BDNF gene in mammals. C_LIO_LIWe observe novel results with 6 sites in our homologous alignment which are under episodic selection in the early regulatory region of BDNF (i.e. the prodomain). C_LIO_LIWe observe 23 pairs of coevolving sites within BDNF. Many of which are a part of complex spatial relationships and are distributed across the entire BDNF gene. C_LIO_LIThese data define exactly how "BDNF is highly conserved" by defining exactly where and how the mammalian BDNF has evolved, confirming the widespread belief that the BDNF prodomain is more prone to change than the mature BDNF protein. C_LI

neuroscience↗

bioRxiv2022

Selection analysis identifies unusual clustered mutational changes in Omicron lineage BA.1 that likely impact Spike function

Among the 30 non-synonymous nucleotide substitutions in the Omicron S-gene are 13 that have only rarely been seen in other SARS-CoV-2 sequences. These mutations cluster within three functionally important regions of the S-gene at sites that will likely impact (i) interactions between subunits of the Spike trimer and the predisposition of subunits to shift from down to up configurations, (ii) interactions of Spike with ACE2 receptors, and (iii) the priming of Spike for membrane fusion. We show here that, based on both the rarity of these 13 mutations in intrapatient sequencing reads and patterns of selection at the codon sites where the mutations occur in SARS-CoV-2 and related sarbecoviruses, prior to the emergence of Omicron the mutations would have been predicted to decrease the fitness of any genomes within which they occurred. We further propose that the mutations in each of the three clusters therefore cooperatively interact to both mitigate their individual fitness costs, and adaptively alter the function of Spike. Given the evident epidemic growth advantages of Omicron over all previously known SARS-CoV-2 lineages, it is crucial to determine both how such complex and highly adaptive mutation constellations were assembled within the Omicron S-gene, and why, despite unprecedented global genomic surveillance efforts, the early stages of this assembly process went completely undetected.

evolutionary biology↗

bioRxiv2022

RASCL: Rapid Assessment Of SARS-CoV-2 Clades Through Molecular Sequence Analysis

An important component of efforts to manage the ongoing COVID19 pandemic is the Rapid Assessment of how natural selection contributes to the emergence and proliferation of potentially dangerous SARS-CoV-2 lineages and CLades (RASCL). The RASCL pipeline enables continuous comparative phylogenetics-based selection analyses of rapidly growing clade-focused genome surveillance datasets, such as those produced following the initial detection of potentially dangerous variants. From such datasets RASCL automatically generates down-sampled codon alignments of individual genes/ORFs containing contextualizing background reference sequences, analyzes these with a battery of selection tests, and outputs results as both machine readable JSON files, and interactive notebook-based visualizations. AvailabilityRASCL is available from a dedicated repository at https://github.com/veg/RASCL and as a Galaxy workflow https://usegalaxy.eu/u/hyphy/w/rascl. Existing clade/variant analysis results are available here: https://observablehq.com/@aglucaci/rascl. ContactDr. Sergei L Kosakovsky Pond (spond@temple.edu). Supplementary informationN/A

bioinformatics↗

bioRxiv2020

Extra base hits: widespread empirical support for instantaneous multiple-nucleotide changes

Despite many attempts to introduce evolutionary models that permit substitutions that instantly alter more than one nucleotide in a codon, the prevailing wisdom remains that such changes are rare and generally negligible (or are reflective of non-biological artifacts, such as alignment errors), and codon models continue to posit that only single nucleotide change have non-zero rates. We develop and test a simple hierarchy of codon-substitution models with non-zero evolutionary rates for only one-nucleotide (1H), one- and two-nucleotide (2H), or any (3H) codon substitutions. Using 35,000 empirical alignments, we find widespread statistical support for multiple hits: 58% of alignments prefer models with 2H allowed, and 22% - with 3H allowed. Analyses of simulated data suggest that these results are not likely to be due to simple artifacts such as model misclassification or alignment errors. Further modeling revealed that synonymous codon island jumping among codons encoding serine, especially along short branches, contributes significantly to this 3H signal. While serine codons were prominently involved in multiple-hit substitutions, there were other common exchanges contributing to better model fit. It appears that a small subset of sites in most alignments have unusual evolutionary dynamics not well explained by existing model formalisms, and that commonly estimated quantities, such as dN/dS ratios may be biased by model misspecification. Our findings highlight the need for continued evaluation of assumptions underlying workhorse evolutionary models and subsequent evolutionary inference techniques. We provide a software implementation for evolutionary biologists to assess the potential impact of extra base hits in their data in the HyPhy package.

bioinformatics↗