Nikolaos Vakirlis
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Nikolaos Vakirlis has a total of 87 co-authors. Here are the top co-authors by number of shared publications:
- Anne Kupczok
Bioinformatics Group, Wageningen University & Research, Wageningen, Netherlands
8 shared publications - Aoife McLysaght
Smurfit Institute of Genetics, Trinity College Dublin, Dublin, Ireland
6 shared publications - Anne-Ruxandra Carvunis
Department of Computational and Systems Biology, School of Medicine, University of Pittsburgh, Pittsburgh, PA 15213, USA
4 shared publications - Christoforos Nikolaou
Biomedical Sciences Research Center "Alexander Fleming
4 shared publications Emilios Tassios
Hellenic Pasteur Institute
4 shared publications- Claudio Casola
Department of Ecology and Conservation Biology, Texas A&M University, College Station, TX, USA
3 shared publications
Latest Publications
- Open Access
Evolutionary origins of protein novelty across an entire yeast subphylum
Novel protein-coding sequences fuel molecular and cellular evolutionary innovations and frequently contribute to species-defining characteristics. They can originate either de novo from previously noncoding sequences or through extreme divergence of already coding ones. How frequently each mechanism occurs and how they shape the structural and functional potential of the resulting proteins remains unclear. Here, we conducted a broad computational investigation of genetic and protein novelty throughout the entire subphylum of Saccharomycotina yeasts. We detected more than 5,000 robust de novo genes across 332 species and compared them to more than 6,000 novel genes resulting from extreme sequence divergence, revealing two quantitatively similar but qualitatively distinct modes of evolution of novelty. A remarkable 40% of de novo proteins are predicted to localize to mitochondria compared to only 20% of divergent, with the latter also being substantially longer and more disordered. A detailed analysis of conservatively predicted tertiary structures of novel proteins shows that ''invention'' of new folds occurs more frequently through de novo emergence. We also illustrate cases of evolutionary ''re-invention'' of existing protein folds from noncoding sequences. Our work deepens our understanding of the origins and importance of novel proteins, opening new directions for further structural and functional characterization.
Open AccessEvolutionary origins of protein novelty across an entire yeast subphylum
Novel protein-coding sequences fuel molecular and cellular evolutionary innovations and frequently contribute to species-defining characteristics. They can originate either de novo from previously noncoding sequences or through extreme divergence of already coding ones. How frequently each mechanism occurs and how they shape the structural and functional potential of the resulting proteins remains unclear. Here, we conducted a broad computational investigation of genetic and protein novelty throughout the entire subphylum of Saccharomycotina yeasts. We detected more than 5,000 robust de novo genes across 332 species and compared them to more than 6,000 novel genes resulting from extreme sequence divergence, revealing two quantitatively similar but qualitatively distinct modes of evolution of novelty. A remarkable 40% of de novo proteins are predicted to localize to mitochondria compared to only 20% of divergent, with the latter also being substantially longer and more disordered. A detailed analysis of conservatively predicted tertiary structures of novel proteins shows that ''invention'' of new folds occurs more frequently through de novo emergence. We also illustrate cases of evolutionary ''re-invention'' of existing protein folds from noncoding sequences. Our work deepens our understanding of the origins and importance of novel proteins, opening new directions for further structural and functional characterization.
- Open Access
- 24 August 2026
- Published by Oxford University Press (OUP) in Genome Biology and Evolution
- Vol. 18 (9)
Identification and Classification of Expressed Orphan Genes, Spurious Orphan Genes, and Conserved Genes in the Human Gut Microbiome
Orphan genes (OGs)—genes lacking detectable homologs outside a species—are widespread in microbial genomes and are thought to contribute to their adaptation and molecular innovation. However, not all predicted OGs may represent novel functional coding sequences. False positive OGs, also called spurious OGs, can arise from gene prediction errors. We reason that OGs lacking detectable expression are more likely to be spurious. To test this, we combined large-scale metatranscriptomic profiling of the human gut microbiome with machine learning to distinguish expressed OGs from spurious ones and compare them with conserved genes (CGs) found in multiple species. Using nearly 5,000 metatranscriptome libraries, we identified ∼218,000 OGs supported by expression evidence, while ∼330,000 predicted OGs lacked detectable expression and were classified as spurious. We extracted 154 features for sequence, structural, and evolutionary properties for each gene and trained XGBoost classifiers while accounting for genomic representation. The models achieved an area under the receiver operating characteristic curve (AUC) of 0.82 in distinguishing expressed OGs from spurious OGs and an AUC of 0.93 in distinguishing expressed OGs from CGs. Interpretation based on SHAP (SHapley Additive exPlanations) revealed clear biological signals. Particularly, expressed orphans were present in more genomes than spurious ones, and expressed OGs were shorter than CGs. This work improves OG discovery and suggests that expressed OGs differ systematically from CGs and spurious OGs in sequence composition, structural constraints, and evolutionary signals.
- 24 August 2026
- Published by Oxford University Press (OUP) in Genome Biology and Evolution
- Vol. 18 (9)
Open AccessIdentification and Classification of Expressed Orphan Genes, Spurious Orphan Genes, and Conserved Genes in the Human Gut Microbiome
Orphan genes (OGs)—genes lacking detectable homologs outside a species—are widespread in microbial genomes and are thought to contribute to their adaptation and molecular innovation. However, not all predicted OGs may represent novel functional coding sequences. False positive OGs, also called spurious OGs, can arise from gene prediction errors. We reason that OGs lacking detectable expression are more likely to be spurious. To test this, we combined large-scale metatranscriptomic profiling of the human gut microbiome with machine learning to distinguish expressed OGs from spurious ones and compare them with conserved genes (CGs) found in multiple species. Using nearly 5,000 metatranscriptome libraries, we identified ∼218,000 OGs supported by expression evidence, while ∼330,000 predicted OGs lacked detectable expression and were classified as spurious. We extracted 154 features for sequence, structural, and evolutionary properties for each gene and trained XGBoost classifiers while accounting for genomic representation. The models achieved an area under the receiver operating characteristic curve (AUC) of 0.82 in distinguishing expressed OGs from spurious OGs and an AUC of 0.93 in distinguishing expressed OGs from CGs. Interpretation based on SHAP (SHapley Additive exPlanations) revealed clear biological signals. Particularly, expressed orphans were present in more genomes than spurious ones, and expressed OGs were shorter than CGs. This work improves OG discovery and suggests that expressed OGs differ systematically from CGs and spurious OGs in sequence composition, structural constraints, and evolutionary signals.
- 2 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Evolution of gene repertoires and new genes in yeasts
Evolution des répertoires de gènes et nouveaux gènes chez les levures: Les répertoires de gènes sont des objets extrêmement dynamiques : Des gènes sont dupliqués et perdus, transférés d’un génome à l’autre et des nouveaux gènes sont créés. L’étude de ces processus et de leur impact sur l’évolution des répertoires de gènes est fondamentale pour notre compréhension de l’énorme diversité de la vie sur terre. J’ai reconstruit les familles des gènes homologues chez les levures du clade Lachancea et je les ai classées en trois catégories selon leur présence chez les espèces en dehors du clade en: transmises verticalement (98.2 %), transmises horizontalement (0.15 %) et spécifiques aux Lachancea (1.63 %). Ensuite, j’ai reconstruit l’évolution de chaque famille de gènes le long de l’arbre phylogénétique des Lachancea en terme de gains et de pertes depuis l’origine du clade. Mes résultats suggèrent que les réarrangements chromosomiques balancés (translocations, inversions) peuvent interrompre, au niveau de leurs points de cassure, la séquence codante des gènes, et entraîner jusqu’à 14 % des pertes de gènes observées (rupture de gène). En outre, j’ai observé des corrélations entre le taux de divergence des séquences codant pour des protéines et les taux de duplication de gènes, de translocations et d’inversions, et de rupture de gène, suggérant l’existence d’une horloge génomique qui coordonnerait ces processus. Par la suite, je me suis focalisé sur l’émergence de nouveaux gènes de novo à partir de séquences non-codantes, dont l’impact global sur les génomes n’est pas encore connu. J’ai pour cela analysé les gènes taxonomiquement restreints aux levures des clades Lachancea et Saccharomyces sensu stricto et j’ai pu identifier un ensemble de 596 gènes ayant fort probablement émergé de novo. Le taux d’émergence de novo est constant chez les levures au sein du même clade mais varie d’un ordre de grandeur entre les 2 clades (2.8 gènes/ma chez les Saccharomyces et 0.27 gènes/ma chez les Lachancea). Ces nouveaux gènes sont distribués uniformément sur les chromosomes. Ils sont le plus souvent orientés de façon divergente par rapport à leur voisin en 5’, ce qui suggère que leur transcription pourrait être initiée au niveau de promoteurs divergents, favorisant ainsi la transition d’une séquence intergénique non transcrite à une séquence codante transcrite (puis traduite). Enfin, j’ai montré que dans certains cas, seul un petit nombre de mutations permettent la création d’un gène bien adapté à son environnement génomique, en comparaison avec des gènes plus «anciens». Cela signifie que sous certaines conditions la transition d’une séquence non-codante vers une séquence codante peut être relativement rapide. Globalement, mes résultats suggèrent que l’émergence de novo est un processus évolutif non négligeable, représentant une source importante de création de nouvelles protéines.
- 2 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Evolution of gene repertoires and new genes in yeasts
Evolution des répertoires de gènes et nouveaux gènes chez les levures: Les répertoires de gènes sont des objets extrêmement dynamiques : Des gènes sont dupliqués et perdus, transférés d’un génome à l’autre et des nouveaux gènes sont créés. L’étude de ces processus et de leur impact sur l’évolution des répertoires de gènes est fondamentale pour notre compréhension de l’énorme diversité de la vie sur terre. J’ai reconstruit les familles des gènes homologues chez les levures du clade Lachancea et je les ai classées en trois catégories selon leur présence chez les espèces en dehors du clade en: transmises verticalement (98.2 %), transmises horizontalement (0.15 %) et spécifiques aux Lachancea (1.63 %). Ensuite, j’ai reconstruit l’évolution de chaque famille de gènes le long de l’arbre phylogénétique des Lachancea en terme de gains et de pertes depuis l’origine du clade. Mes résultats suggèrent que les réarrangements chromosomiques balancés (translocations, inversions) peuvent interrompre, au niveau de leurs points de cassure, la séquence codante des gènes, et entraîner jusqu’à 14 % des pertes de gènes observées (rupture de gène). En outre, j’ai observé des corrélations entre le taux de divergence des séquences codant pour des protéines et les taux de duplication de gènes, de translocations et d’inversions, et de rupture de gène, suggérant l’existence d’une horloge génomique qui coordonnerait ces processus. Par la suite, je me suis focalisé sur l’émergence de nouveaux gènes de novo à partir de séquences non-codantes, dont l’impact global sur les génomes n’est pas encore connu. J’ai pour cela analysé les gènes taxonomiquement restreints aux levures des clades Lachancea et Saccharomyces sensu stricto et j’ai pu identifier un ensemble de 596 gènes ayant fort probablement émergé de novo. Le taux d’émergence de novo est constant chez les levures au sein du même clade mais varie d’un ordre de grandeur entre les 2 clades (2.8 gènes/ma chez les Saccharomyces et 0.27 gènes/ma chez les Lachancea). Ces nouveaux gènes sont distribués uniformément sur les chromosomes. Ils sont le plus souvent orientés de façon divergente par rapport à leur voisin en 5’, ce qui suggère que leur transcription pourrait être initiée au niveau de promoteurs divergents, favorisant ainsi la transition d’une séquence intergénique non transcrite à une séquence codante transcrite (puis traduite). Enfin, j’ai montré que dans certains cas, seul un petit nombre de mutations permettent la création d’un gène bien adapté à son environnement génomique, en comparaison avec des gènes plus «anciens». Cela signifie que sous certaines conditions la transition d’une séquence non-codante vers une séquence codante peut être relativement rapide. Globalement, mes résultats suggèrent que l’émergence de novo est un processus évolutif non négligeable, représentant une source importante de création de nouvelles protéines.
- Top Cited PapersOpen Access
- 19 January 2026
- Published by Oxford University Press (OUP) in Genome Biology and Evolution
- Vol. 18 (2)
Degradation Determinants Are Abundant in Human Noncanonical Proteins and Minor Annotated Isoforms
The comprehensive characterization of human proteins, a key objective in contemporary biology, has been revolutionized by the identification of thousands of potential novel proteins through ribosome profiling and proteomics. Determining the physiological activity of these noncanonical proteins has proven difficult, because they are encoded by different types of coding regions and tend to share no sequence similarity with canonical polypeptides. Evidence from immunopeptidomic assays combined with a better understanding of the quality control of protein synthesis suggest that many noncanonical proteins may possess low stability in the cellular environment. Here, we tested this hypothesis by analyzing the frequency of multiple sequence features eliciting either proteasomal degradation or autophagy across 91,003 canonical (annotated) protein isoforms and 11,499 noncanonical proteins. Overall, noncanonical proteins were enriched for degradation-related features compared to all canonical proteins. Notably, degradation determinants were also enriched in canonical protein isoforms starting with a non-methionine amino acid. Analyses of original and shuffled sequences showed evidence of selective pressure either against or toward the accumulation of specific degradation signatures only in major isoforms of canonical proteins. However, stability was significantly higher in noncanonical proteins with evidence of phenotypic effects upon knock-out in cell lines. Notably, we found that the C-terminal tail hydrophobicity represents a reliable proxy for degradation propensity with potential applications in identifying functional noncanonical proteins. These findings underscore the critical role of degradation processes in regulating the half-life of noncanonical proteins and demonstrate the power of degradation-associated signatures in discriminating noncanonical genes likely to encode biologically functional molecules.
- 19 January 2026
- Published by Oxford University Press (OUP) in Genome Biology and Evolution
- Vol. 18 (2)
Top Cited PapersOpen AccessDegradation Determinants Are Abundant in Human Noncanonical Proteins and Minor Annotated Isoforms
The comprehensive characterization of human proteins, a key objective in contemporary biology, has been revolutionized by the identification of thousands of potential novel proteins through ribosome profiling and proteomics. Determining the physiological activity of these noncanonical proteins has proven difficult, because they are encoded by different types of coding regions and tend to share no sequence similarity with canonical polypeptides. Evidence from immunopeptidomic assays combined with a better understanding of the quality control of protein synthesis suggest that many noncanonical proteins may possess low stability in the cellular environment. Here, we tested this hypothesis by analyzing the frequency of multiple sequence features eliciting either proteasomal degradation or autophagy across 91,003 canonical (annotated) protein isoforms and 11,499 noncanonical proteins. Overall, noncanonical proteins were enriched for degradation-related features compared to all canonical proteins. Notably, degradation determinants were also enriched in canonical protein isoforms starting with a non-methionine amino acid. Analyses of original and shuffled sequences showed evidence of selective pressure either against or toward the accumulation of specific degradation signatures only in major isoforms of canonical proteins. However, stability was significantly higher in noncanonical proteins with evidence of phenotypic effects upon knock-out in cell lines. Notably, we found that the C-terminal tail hydrophobicity represents a reliable proxy for degradation propensity with potential applications in identifying functional noncanonical proteins. These findings underscore the critical role of degradation processes in regulating the half-life of noncanonical proteins and demonstrate the power of degradation-associated signatures in discriminating noncanonical genes likely to encode biologically functional molecules.
- Open Access
Identification and Classification of Expressed Orphan Genes, Spurious Orphan Genes, and Conserved Genes in the Human Gut Microbiome
Orphan genes - genes lacking detectable homologs outside a species - are widespread in microbial genomes and are thought to contribute to their adaptation and molecular innovation. However, not all predicted orphan genes may represent novel functional coding sequences. False positive orphan genes, also called spurious orphan genes, can arise from gene prediction errors. We reason that orphan genes lacking detectable expression are more likely to be spurious. To this end, we combined large-scale metatranscriptomic profiling of the human gut microbiome with machine learning to distinguish expressed orphan genes from spurious ones and to compare them with conserved genes found in multiple species. Using nearly 5,000 metatranscriptome libraries, we identified ∼218,000 orphan genes supported by expression evidence, while ∼330,000 predicted orphan genes lacked detectable expression, and were classified as spurious. We extracted 154 sequence, structural, and evolutionary features for each gene and trained XGBoost classifiers while accounting for genomic representation. The models achieved an area under the receiver operating characteristic curve (AUC) of 0.82 in distinguishing expressed orphan genes from spurious orphan genes and 0.93 in distinguishing expressed orphan genes from conserved genes. SHAP-based interpretation revealed clear biological signals. E.g., expressed orphans were present in more genomes than spurious ones and expressed orphan genes were shorter than conserved genes. This work improves orphan gene discovery and suggests that expressed orphan genes differ systematically from conserved genes and spurious orphan genes in sequence composition, structural constraints, and evolutionary signals.
Open AccessIdentification and Classification of Expressed Orphan Genes, Spurious Orphan Genes, and Conserved Genes in the Human Gut Microbiome
Orphan genes - genes lacking detectable homologs outside a species - are widespread in microbial genomes and are thought to contribute to their adaptation and molecular innovation. However, not all predicted orphan genes may represent novel functional coding sequences. False positive orphan genes, also called spurious orphan genes, can arise from gene prediction errors. We reason that orphan genes lacking detectable expression are more likely to be spurious. To this end, we combined large-scale metatranscriptomic profiling of the human gut microbiome with machine learning to distinguish expressed orphan genes from spurious ones and to compare them with conserved genes found in multiple species. Using nearly 5,000 metatranscriptome libraries, we identified ∼218,000 orphan genes supported by expression evidence, while ∼330,000 predicted orphan genes lacked detectable expression, and were classified as spurious. We extracted 154 sequence, structural, and evolutionary features for each gene and trained XGBoost classifiers while accounting for genomic representation. The models achieved an area under the receiver operating characteristic curve (AUC) of 0.82 in distinguishing expressed orphan genes from spurious orphan genes and 0.93 in distinguishing expressed orphan genes from conserved genes. SHAP-based interpretation revealed clear biological signals. E.g., expressed orphans were present in more genomes than spurious ones and expressed orphan genes were shorter than conserved genes. This work improves orphan gene discovery and suggests that expressed orphan genes differ systematically from conserved genes and spurious orphan genes in sequence composition, structural constraints, and evolutionary signals.
- Open Access
- 27 December 2025
- Published by Oxford University Press (OUP) in Bioinformatics Advances
- Vol. 6 (1)
Machine learning can distinguish orphans that have resulted from sequence divergence beyond recognition
Motivation: Species-specific orphan genes lack homologues outside of a given taxon and frequently underlie unique species traits. Orphans can result from sequence divergence beyond recognition, when homologous proteins diverge to an extent at which sequence similarity search algorithms can no longer identify them as homologues, but they can also evolve de novo from previously noncoding sequences, in which case homologous protein-coding genes truly do not exist. Results: Here we propose that sequence divergent orphans might be recognizable from their patterns of non-statistically significant similarity hits which are typically discarded. To test this, we simulated diverged orphan protein sequences under varying parameters. Using reversed protein sequences as negative control, we trained machine learning classifiers on features extracted from similarity search output. We found that this approach works, but performance of the models depends on the simulation parameters, with ∼90% accuracy when the underlying simulated divergence was moderate and ∼70% when it is extreme. When applying our classifiers on a set of real orphans we found that ∼30% of them are predicted to be divergent and these are shorter and more disordered than the rest. Our work contributes to the effort of better understanding how genetic novelty arises. Availability and implementation: The models and data used can be found at https://github.com/emiliostassios/Classification-of-divergent-genes-using-ML
- 27 December 2025
- Published by Oxford University Press (OUP) in Bioinformatics Advances
- Vol. 6 (1)
Open AccessMachine learning can distinguish orphans that have resulted from sequence divergence beyond recognition
Motivation: Species-specific orphan genes lack homologues outside of a given taxon and frequently underlie unique species traits. Orphans can result from sequence divergence beyond recognition, when homologous proteins diverge to an extent at which sequence similarity search algorithms can no longer identify them as homologues, but they can also evolve de novo from previously noncoding sequences, in which case homologous protein-coding genes truly do not exist. Results: Here we propose that sequence divergent orphans might be recognizable from their patterns of non-statistically significant similarity hits which are typically discarded. To test this, we simulated diverged orphan protein sequences under varying parameters. Using reversed protein sequences as negative control, we trained machine learning classifiers on features extracted from similarity search output. We found that this approach works, but performance of the models depends on the simulation parameters, with ∼90% accuracy when the underlying simulated divergence was moderate and ∼70% when it is extreme. When applying our classifiers on a set of real orphans we found that ∼30% of them are predicted to be divergent and these are shorter and more disordered than the rest. Our work contributes to the effort of better understanding how genetic novelty arises. Availability and implementation: The models and data used can be found at https://github.com/emiliostassios/Classification-of-divergent-genes-using-ML
- Open Access
- 28 November 2025
- Published by Oxford University Press (OUP) in Genome Biology and Evolution
- Vol. 17 (12)
De Novo Genes: Current Status and Future Goals
The recent Society for Molecular Biology and Evolution Satellite Meeting on De Novo Gene Birth, hosted at Texas A&M University on November 6 to 9, 2023, represented the first-ever opportunity for scientists studying the evolution and biology of de novo genes to gather through a dedicated meeting and discuss about groundbreaking discoveries in this emerging and exciting field of gene evolution. In this perspective, we discuss recent advances and major open questions in de novo gene emergence and evolution that were presented at the SMBE satellite meeting, as well as some of the key recent findings published before or since the conference. These key themes include de novo gene identification, function, and evolution, what we are learning about de novo genes from experimental analyses of random peptides, de novo gene birth and microproteins, and the role of de novo genes in human disease.
- 28 November 2025
- Published by Oxford University Press (OUP) in Genome Biology and Evolution
- Vol. 17 (12)
Open AccessDe Novo Genes: Current Status and Future Goals
The recent Society for Molecular Biology and Evolution Satellite Meeting on De Novo Gene Birth, hosted at Texas A&M University on November 6 to 9, 2023, represented the first-ever opportunity for scientists studying the evolution and biology of de novo genes to gather through a dedicated meeting and discuss about groundbreaking discoveries in this emerging and exciting field of gene evolution. In this perspective, we discuss recent advances and major open questions in de novo gene emergence and evolution that were presented at the SMBE satellite meeting, as well as some of the key recent findings published before or since the conference. These key themes include de novo gene identification, function, and evolution, what we are learning about de novo genes from experimental analyses of random peptides, de novo gene birth and microproteins, and the role of de novo genes in human disease.
- Open Access
- 12 July 2025
- Published by Oxford University Press (OUP) in Journal of Evolutionary Biology
- Vol. 38 (9) , 1272-1277
Intergenic polyA/T tracts explain the propensity of yeast de novo genes to encode transmembrane domains
New genes can emerge de novo from non-genic genomic regions. In budding yeast, computational predictions have shown that intergenic regions harbour a higher-than-expected propensity to encode transmembrane domains, if theoretically translated into proteins. This propensity seems to be linked to the high prevalence of predicted transmembrane domains in evolutionarily young genes. However, what accounts for this enriched propensity is not known. Here, we show that specific arrangements of polyA/T tracts, which are abundant and enriched in yeast intergenic regions, explain this observation. These tracts are known to function as nucleosome-depleted regions, which prevent or reduce nucleosome formation to enable transcription of surrounding genes. We provide evidence that these polyA/T tracts have been repeatedly coopted through de novo gene emergence for the evolution of novel small genes encoding proteins with predicted transmembrane domains. These findings support a previously proposed “transmembrane-first” model of de novo gene birth and help explain why evolutionarily young yeast genes are rich in transmembrane domains. They contribute to our understanding of the process of de novo gene evolution and show how seemingly distinct but potentially interacting levels of functionality can exist within the same genomic loci.
- 12 July 2025
- Published by Oxford University Press (OUP) in Journal of Evolutionary Biology
- Vol. 38 (9) , 1272-1277
Open AccessIntergenic polyA/T tracts explain the propensity of yeast de novo genes to encode transmembrane domains
New genes can emerge de novo from non-genic genomic regions. In budding yeast, computational predictions have shown that intergenic regions harbour a higher-than-expected propensity to encode transmembrane domains, if theoretically translated into proteins. This propensity seems to be linked to the high prevalence of predicted transmembrane domains in evolutionarily young genes. However, what accounts for this enriched propensity is not known. Here, we show that specific arrangements of polyA/T tracts, which are abundant and enriched in yeast intergenic regions, explain this observation. These tracts are known to function as nucleosome-depleted regions, which prevent or reduce nucleosome formation to enable transcription of surrounding genes. We provide evidence that these polyA/T tracts have been repeatedly coopted through de novo gene emergence for the evolution of novel small genes encoding proteins with predicted transmembrane domains. These findings support a previously proposed “transmembrane-first” model of de novo gene birth and help explain why evolutionarily young yeast genes are rich in transmembrane domains. They contribute to our understanding of the process of de novo gene evolution and show how seemingly distinct but potentially interacting levels of functionality can exist within the same genomic loci.
- Open Access
Machine learning can identify orphans that have diverged into the “twilight zone” of sequence similarity
Species-specific orphan genes (orphans) lack homologues outside of a given taxon and frequently underlie unique species traits. It is thus important to elucidate the evolutionary origins of orphans. Orphan genes can result from sequence divergence beyond recognition, when homologous proteins diverge to an extent at which tools that rely on sequence similarity to establish homology can no longer identify them as homologues. Orphans can also result from other processes, including de novo gene emergence from previously noncoding sequences, in which a homologous protein-coding gene truly does not exist. Here we propose that orphans resulting from divergence might be recognizable from their patterns of non-statistically significant similarity hits which are almost always discarded. To test this, we simulated diverged orphan protein sequences based on conserved proteins from the Unified Human Gastrointestinal Protein catalogue (UHGP) and used reversed protein sequences as negative data sets. We trained four machine learning classifiers on features extracted from the similarity search tool DIAMOND’s output, like total query coverage or maximum bit score. We tested the influence of evolutionary parameters such as simulation tree branch length, indel rate and among-site rate heterogeneity. We found that the performance of the models depends on the simulation parameters: when the underlying simulated divergence was moderate, accuracy reached ∼90%, but when extremely diverged scenarios where simulated accuracy dropped to ∼70%. The most important features for the classification were the number of alignments (hits) and the minimum hit E-value. When applying our classifier on a set of ∼170,000 eligible simulated orphans from the UHGP dataset, we found that ∼30% of them are predicted to be divergent and these are shorter and more disordered than the rest. Our classifiers and pre-processing python scripts are available at https://github.com/emiliostassios/Classification-of-divergent-genes-using-ML and can be readily used as a computationally fast means to obtain a candidate set of diverged orphans from any similarity search output. Therefore, our work allows to study such orphans across the tree of life and in doing so to recover cases of remote homology, get better estimates of the evolutionary age of protein families, detect cases of rapid divergence and more generally better understand how genetic novelty arises.
Open AccessMachine learning can identify orphans that have diverged into the “twilight zone” of sequence similarity
Species-specific orphan genes (orphans) lack homologues outside of a given taxon and frequently underlie unique species traits. It is thus important to elucidate the evolutionary origins of orphans. Orphan genes can result from sequence divergence beyond recognition, when homologous proteins diverge to an extent at which tools that rely on sequence similarity to establish homology can no longer identify them as homologues. Orphans can also result from other processes, including de novo gene emergence from previously noncoding sequences, in which a homologous protein-coding gene truly does not exist. Here we propose that orphans resulting from divergence might be recognizable from their patterns of non-statistically significant similarity hits which are almost always discarded. To test this, we simulated diverged orphan protein sequences based on conserved proteins from the Unified Human Gastrointestinal Protein catalogue (UHGP) and used reversed protein sequences as negative data sets. We trained four machine learning classifiers on features extracted from the similarity search tool DIAMOND’s output, like total query coverage or maximum bit score. We tested the influence of evolutionary parameters such as simulation tree branch length, indel rate and among-site rate heterogeneity. We found that the performance of the models depends on the simulation parameters: when the underlying simulated divergence was moderate, accuracy reached ∼90%, but when extremely diverged scenarios where simulated accuracy dropped to ∼70%. The most important features for the classification were the number of alignments (hits) and the minimum hit E-value. When applying our classifier on a set of ∼170,000 eligible simulated orphans from the UHGP dataset, we found that ∼30% of them are predicted to be divergent and these are shorter and more disordered than the rest. Our classifiers and pre-processing python scripts are available at https://github.com/emiliostassios/Classification-of-divergent-genes-using-ML and can be readily used as a computationally fast means to obtain a candidate set of diverged orphans from any similarity search output. Therefore, our work allows to study such orphans across the tree of life and in doing so to recover cases of remote homology, get better estimates of the evolutionary age of protein families, detect cases of rapid divergence and more generally better understand how genetic novelty arises.
- Open Access
Intergenic polyA/T tracts explain the propensity of yeast de novo genes to encode transmembrane domains
New genes can emerge de novo from non-genic genomic regions. In budding yeast, computational predictions have shown that intergenic regions harbor a higher-than-expected propensity to encode transmembrane domains, and this propensity seems to be linked to the high prevalence of predicted transmembrane domains in evolutionarily young genes. However, what accounts for this enriched propensity is not known. Here we show that specific arrangements of polyA/T tracts, which are abundant and enriched in yeast intergenic regions explain this observation. We provide evidence that polyA/T tracts, which are known to act as Nucleosome Depleted Regions in a regulatory context, have been coopted through de novo gene emergence for the evolution of novel small genes encoding proteins with predicted transmembrane domains. Our findings contribute to our understanding of the process of de novo gene evolution and show how seemingly distinct but interacting levels of functionality can exist within the same genomic loci.
Open AccessIntergenic polyA/T tracts explain the propensity of yeast de novo genes to encode transmembrane domains
New genes can emerge de novo from non-genic genomic regions. In budding yeast, computational predictions have shown that intergenic regions harbor a higher-than-expected propensity to encode transmembrane domains, and this propensity seems to be linked to the high prevalence of predicted transmembrane domains in evolutionarily young genes. However, what accounts for this enriched propensity is not known. Here we show that specific arrangements of polyA/T tracts, which are abundant and enriched in yeast intergenic regions explain this observation. We provide evidence that polyA/T tracts, which are known to act as Nucleosome Depleted Regions in a regulatory context, have been coopted through de novo gene emergence for the evolution of novel small genes encoding proteins with predicted transmembrane domains. Our findings contribute to our understanding of the process of de novo gene evolution and show how seemingly distinct but interacting levels of functionality can exist within the same genomic loci.