Biology subjects

Castillo, S.

Publications and source records attributed to Castillo, S..

15 recordsLinked to original sources

bioRxiv2026

ERK overstimulation leads to cell hyperproliferation in hereditary hemorrhagic telangiectasia landscape

Hereditary hemorrhagic telangiectasia (HHT) is a rare vascular disorder caused by pathogenic variants in members of the BMP9/ALK1 signaling hub. In the present study we show that, regardless of whether the alterations are caused by reduced BMP9/ALK1 signaling (pathogenic variants in the ENG or ALK1 genes) or by overactivation of this pathway (such as the SMAD6 pathogenic variants), all are associated with increased endothelial cell (EC) proliferation and high levels of ERK MAPK activation in patient biopsies. We reproduced this phenotype in vitro in ECs lacking SMAD6 or after SMAD1 knockdown using siRNA. Loss of SMAD6 leads to dysregulation of the Notch pathway, with downregulation of phosphatases and consequent overstimulation of ERK. In normal ECs, BMP9 and Notch signaling inhibit ERK activity by upregulating PPP1R3C, a regulatory subunit of the PP1 phosphatase. Notably, BMP9-mediated inhibition of ERK is abolished when cells are transfected with siRNA targeting PPP1R3C. ERK hyperactivation was also observed in an HHT2 mouse model (ALK1-2loxP;Cdh5-CreERT2). Loss of both ALK1 alleles in adult mice leads to vascular failure and hemorrhages in the lung and intestine; these injuries are significantly reduced by treatment with the MEK/ERK inhibitor selumetinib. Overall, our work identifies a key role for ERK activation involved in HHT pathogenesis, suggesting that ERK inhibition may represent a promising therapeutic strategy for these patients. Translational PerspectiveHereditary hemorrhagic telangiectasia (HHTs) is a rare vascular disorder caused by mutations in members of the BMP9/ALK1 signaling hub. In the present study we show that all different forms of HHTs are associated with increased endothelial cell (EC) proliferation, which correlates with high levels of ERK activation in patient biopsies. ERK hyperactivation is also observed in an HHT2 mouse model in which loss of both ALK1 alleles in adult mice leads to vascular failure and hemorrhages in the lung and intestine. These injuries are significantly reduced by treatment with the MEK/ERK inhibitor selumetinib. Overall, our work identifies a key role for ERK activation in HHT pathogenesis, suggesting that ERK/MEK inhibitors may represent a promising therapeutic strategy for these patients.

pathology↗

bioRxiv2026

Comp2GPR: A Sequence-Driven Framework for Gene.Protein-Reaction Rule Reconstruction

Accurate gene-protein-reaction (GPR) associations are essential for the predictive performance of genome-scale metabolic models (GEMs), as they define the mapping between genes, enzymes, and metabolic reactions. However, GPR rules are often incomplete or inconsistent due to limitations in annotation transfer and the ambiguous representation of multi-subunit protein complexes, leading to errors in downstream analyses such as gene essentiality prediction. Here, I introduce Comp2GPR, an automated pipeline for reconstructing GPR rules that integrates curated protein complex information with sequence-level evidence. Protein complexes were sourced from the Complex Portal and subjected to an AI-assisted curation workflow to retain only metabolically relevant assemblies. Comp2GPR combines deterministic sequence similarity mapping with explicit rule construction to generate Boolean GPR expressions that accurately represent obligate subunit relationships and isoenzyme redundancy. I evaluated the impact of the reconstructed GPR rules by integrating them into the Yeast9 metabolic model and comparing gene essentiality predictions with the original model. While global performance metrics remained largely unchanged, the updated model achieved a net improvement in prediction accuracy through gene-level corrections. Overall, Comp2GPR demonstrates that combining curated protein complex data with sequence-based validation improves the accuracy, interpretability, and reproducibility of GPR rules. The method provides a robust framework for enhancing metabolic model annotations and supports more reliable simulation-based analyses.

bioinformatics↗

bioRxiv2026

ActSeekN: A Structural-Motif-Based Pipeline for Interpretable Enzyme Function Annotation

Accurate enzyme function annotation remains a major bottleneck in genome analysis despite the rapid expansion of available protein sequence and structure data. Most existing methods rely on sequence similarity or machine-learning representations, which often perform poorly for proteins with low sequence identity or convergent evolutionary histories. Because enzymatic activity is determined by the three-dimensional arrangement of catalytic and binding-site residues, structure-based approaches offer a mechanistically grounded alternative. However, their broader application has been constrained by the limited size and coverage of curated active-site reference databases. To address this challenge, we developed ActSeekN, a structural-motif-based functional annotation pipeline that combines the ActSeek active-site search algorithm with a newly constructed large-scale reference database derived from AlphaFold-predicted structures, UniProt annotations, and curated catalytic residue information. This framework enables rapid and scalable identification of conserved catalytic motifs across structurally related proteins, allowing function to be transferred on the basis of local three-dimensional catalytic geometry rather than global sequence similarity. In this way, ActSeekN overcomes a central limitation of previous structure-based methods by expanding the searchable space of catalytic motifs while retaining mechanistic interpretability. Benchmarking against state-of-the-art machine-learning approaches demonstrates competitive or superior performance. Applications to yeast, human, and Trichoderma reesei proteomes refine existing annotations, complete partial EC assignments, and identify previously unrecognized enzymatic functions, highlighting ActSeekN as a powerful tool for genome annotation and biotechnology. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=99 SRC="FIGDIR/small/720574v1_ufig1.gif" ALT="Figure 1"> View larger version (37K): org.highwire.dtl.DTLVardef@f5da16org.highwire.dtl.DTLVardef@c0faa4org.highwire.dtl.DTLVardef@18765fforg.highwire.dtl.DTLVardef@3960b9_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2025

HVRLocator: A Computationally Efficient Tool for Identifying Hypervariable Regions in 16S rRNA Big Datasets

BackgroundAmplicon sequencing of the 16S rRNA gene is widely used to assess microbial diversity due to its cost-effectiveness and efficiency. However, public 16S rRNA datasets often lack standardized metadata, particularly information on the sequenced hypervariable regions or primers used, which are critical for accurate analysis and data reuse. To address this, we present the HVRLocator, a computational tool that reliably identifies sequenced hypervariable regions, enhancing metadata quality and enabling more robust large-scale microbiome studies. ResultsThe HVRLocator tool processed samples at an average rate of 0.147 per minute. Validation confirmed 100% accuracy in predicting alignment positions, correctly matching sequences to the expected primer regions based on literature. We demonstrated how to use the tool to select appropriate and comparable sequences for building a global bacterial database from V4 region amplicons of the 16S rRNA gene. Using HVRLocator, we selected 36,217 valid samples out of 45,882 runs, enabling us to identify cases where metadata incorrectly labeled sequences as targeting the V4 region. ConclusionEven when metadata is available, it can be inaccurate or misleading. HVRLocator offers a reliable and efficient method to identify the exact hypervariable sequenced region, ensuring accurate processing of large-scale 16S rRNA amplicon data. By bypassing inconsistent metadata and literature, it streamlines data curation and enhances the reliability of microbial studies, syntheses, and meta-analyses. Its use is essential for critically evaluating published data and enabling accurate and reproducible research in microbial ecology.

bioinformatics↗

bioRxiv2025

Label-Free Fluorescence Microscopy Reveals Multiphase Organization in Biomolecular Condensates

Phase transitions of proteins and nucleic acids (NA) leading to the formation of biomolecular condensates have been linked to various biological functions. Given the growing number of proteins/NA predicted to undergo liquid-liquid phase separation (LLPS), efficient tools to investigate this behavior are critical to advancing our understanding of biomolecular condensate function. The current standard used to study LLPS involves techniques that utilize exogenous fluorophore labels. The labeling process is often costly and time-consuming and comes with associated complexity that arises from unknown interactions from the bulky fluorescent tags. These aspects limit high throughput analysis of protein/NA phase separation based on external fluorophore labeling. Here, we report the discovery that intrinsic fluorescence, well into the visible spectrum, arises as an emergent property of biomolecular condensates. Leveraging this intrinsic fluorescence, we study condensate formation, directly measure their internal dynamics via Fluorescence Recovery after Photobleaching (FRAP), and examine the 3D morphology and transitions to various multiphase architectures. Through this approach, we find that a variety of G-quadruplex DNA readily form droplets with histone H1 and display dynamic exchange. In addition, we directly demonstrate that the 3D morphology, core-shell architecture, and sub-compartmentalization of condensate droplets are tunable via the charge ratio of components in solution and NA hybridization. Our method utilizes an inherent property of condensates, thus is broadly applicable to any phase-separated systems and can advance our understanding of biological phase transition.

biophysics↗

bioRxiv2025

ActSeek: Fast and accurate search algorithm of active sites in Alphafold database

Finding proteins with specific functions by mining modern databases can potentially lead to substantial advancements in wide range of fields, from medicine and biotechnology to material science. Currently available algorithms enable mining of proteins based on their sequence or structure. However, activities of many proteins, such as enzymes and drug targets, are dictated by active site residues and their surroundings rather than the overall structure or sequence of a protein. Here we present ActSeek - a computer vision-inspired fast program - that searches structural databases for proteins with active sites similar to the seed protein. ActSeek is implemented to mine proteins with desired active site environments from the Alphafold database. The potential of ActSeek to find innovative solutions to the worlds most pressing challenges is demonstrated by finding enzymes that may be used to produce biodegradable plastics or degrade plastics, as well as potential off-targets for common drug molecules.

bioinformatics↗

bioRxiv2024

New-to-nature PHA synthase design using deep learning

Polyhydroxyalkanaoate (PHA) synthases are a group of complex, dimeric enzymes which catalyse polymerization of Rhydroxyacids into PHAs. PHA properties depend on their monomer composition but enzymes found in nature have narrow specificities to certain R-hydroxyacids. In this study, a conditional variational autoencoder was used for the first time to design new-to-nature PHA synthases. The model was trained with natural protein sequences obtained from Uniprot and was used for the creation of approximately 10 000 new PHA synthase enzymes. Out of these, 16 sequences were selected for in vivo validation. The selection criteria included the presence of conserved residues such as catalytic amino acids and amino acids in the dimer interface and structural features like the number of -helixes in the N-terminal part of the enzyme. Two of the new-to-nature PHA synthases that had substantial numbers of amino acid substitutions (87 and 98) with respect to the most similar native enzymes were confirmed active and produced poly(hydroxybutyrate) (PHB) when expressed in yeast S. cerevisiae. PHA including PHB have high potential as biodegradable and biocompatible materials. Ultimately the model-designed new-to-nature PHA synthases, could expand the PHA material properties to suit new application areas.

bioengineering↗

bioRxiv2023

Reconstruction of compartmentalized genome-scalemetabolic models using deep learning for over 800fungi

Eukaryotic metabolism is organized into subcellular compartments enclosed by lipid-membranes. Since most metabolites do not move freely across the membranes, the compartmentalization influences metabolic pathway connectivities. Thus, compartmentalizing the pathways also in genome-scale metabolic models (GEMs) is essential for accurate predictions of eukaryotic cells metabolic phenotypes. Compartmentalization has manually been introduced into the model Eukaryote GEMs like for yeast Saccharomyces cerevisiae. Non-model organisms GEMs can be automatically reconstructed from genome data. However, the existing GEM reconstruction methods do not introduce compartmentalization into the models. To that end, we integrated our novel deep learning protein localization prediction and protein functional annotation into a top-down GEM reconstruction for automatically creating species-specific compartmentalized GEMs. We developed also a universal fungal GEM for top-down reconstruction of species-specifically compartmentalized GEMs and reconstructed models for 834 species. The novel protein localization prediction outperformed the state of the art in classifying proteins into multiple compartments and integrating the enzyme localization predictions to functional annotations improved the top-down model reconstruction. Interestingly, the clustering of the fungal GEMs reconstructed with predicted species-specific enzyme localization resembled more closely the phylogenetic relationships of the species than that of the GEMS without the compartmentalization. Compartmentalization of metabolism in fungi differs from other eukaryotes but the enzyme subcellular localizations vary also among fungal species. The Fungal kingdom encompasses species important for human health, environment, and industrial biotechnology. The reconstructed fungal GEM set offers valuable tools for e.g., predicting metabolic phenotypes of e.g., mushrooms or single-cellular fungi, the roles of eukaryotes in microbial communities, designs optimizing eukaryotic hosts for industrial chemical production, and identifying drug targets against pathogenic fungi. Beyond fungi, the compartmentalized GEM reconstruction method allows combining other universal GEMs for developing cell type specific GEMs for higher eukaryotes including plants, insects, and mammals.

systems biology↗

bioRxiv2023

Genome-scale metabolic models reveal determinants of phenotypic differences in non-Saccharomyces yeasts

BackgroundUse of alternative non-Saccharomyces yeasts in wine and beer brewing has gained more attention the recent years. This is both due to the desire to obtain a wider variety of flavours in the product and to reduce the final alcohol content. Given the metabolic differences between the yeast species, we wanted to account for some of the differences by using in silico models. ResultsWe created and studied genome-scale metabolic models of five different non-Saccharomyces species using an automated processes. These were: Metschnikowia pulcherrima, Lachancea thermotolerans, Hanseniaspora osmophila and Kluyveromyces lactis. Using the models, we predicted that M. pulcherrima, when compared to the other species, conducts more respiration and thus produces less fermentation products, a finding which agrees with experimental data. Complex I of the electron transport chain was to be present in M. pulcherrima, but absent in the others. The predicted importance of Complex I was diminished when we incorporated constraints on the amount of enzymatic protein, as this shifts the metabolism towards fermentation. ConclusionsOur results suggest that Complex I in the electron transport chain is a key differentiator between Metschnikowia pulcherrima and the other yeasts considered. Yet, more annotations and experimental data have the potential to improve model quality in order to increase fidelity and confidence in these results. Further experiments should be conducted to confirm the in vivo effect of Complex I in M. pulcherrima and its respiratory metabolism.

systems biology↗

bioRxiv2023

Chromosome-aware phylogenomics of Assassin Bugs (Hemiptera: Reduvioidea) elucidates ancient gene conflict

Though the phylogenetic signal of loci on sex chromosomes can differ from those on autosomes, chromosomal-level genome assemblies for non-vertebrates are still relatively scarce and conservation of chromosomal gene content across deep phylogenetic scales has therefore remained largely unexplored. We here assemble a uniquely large and diverse set of samples (17 Anchored Hybrid Enrichment [AHE], 24 RNA-Seq, and 70 whole-genome sequencing [WGS] samples of variable depth) for the medically important assassin bugs (Reduvioidea). We assess the performance of genes based on multiple features (e.g., nucleotide vs. amino acid, nuclear vs. mitochondrial, and autosomal vs. X chromosomal) and employ different methods (concatenation and coalescence analyses) to reconstruct the unresolved phylogeny of this diverse ([~]7,000 spp.) and old (>180 MYA) group. Our results show that genes on the X chromosome are more likely to have discordant phylogenies than those on autosomes. We find that the X chromosome conflict is driven by high gene substitution rates that impact accuracy of phylogenetic inference. However, gene tree clustering showed strong conflict even after discounting variable third codon positions. Alternative topologies were not particularly enriched for sex chromosome loci, but spread across the genome. We conclude that binning genes to autosomal or sex chromosomes may result in a more accurate picture of the complex evolutionary history of a clade.

genomics↗

bioRxiv2022

Characterization of Genomic Diversity In Bacteriophages Infecting Rhodococcus

Bacteriophages are globally ubiquitous viruses that infect bacteria. With nearly 4,000 sequenced genomes of phages infecting the phylum Actinobacteria available, genomic analyses of these actinobacteriophage genomes has been instrumental in uncovering a diverse genomic landscape often characterized by genome mosaicism. Here, we describe the genomic characterization of 57 sequenced phages capable of infecting the genus Rhodococcus. These phages were previously isolated at multiple institutions by students in the SEA-PHAGES program using four different species of Rhodococcus. Most Rhodococcus phages have been grouped into 4 clusters based on their genomic similarities; 13 phages are singletons too genetically distinct for clustering. These clusters and singletons contain Siphoviridae and Myoviridae phages, and most contain integrase and repressor genes indicative of a potential lysogenic life cycle. The genome size of these phages varies from 14,270 bp to 142,586 bp and their G+C% content ranges from 41.2-68.4%, while that of their Rhodococcus hosts typically exceeds 60%. Through comparative genomic analyses, it was revealed that these Rhodococcus phages display high intracluster similarity but low intercluster similarity, despite their shared ability to infect the same host genus. Additionally, these Rhodococcus phages share similarities with phages that infect other Actinobacterial hosts such as Gordonia, Streptomyces and Arthrobacter.

microbiology↗

bioRxiv2022

A Canine Model of Chronic Ischemic Heart Failure

BackgroundThere is a paucity of novel therapeutics for chronic heart failure (HF). Preclinical large animal models of chronic HF are crucial to translating fundamental discoveries. Canine models of ischemic cardiomyopathy are frequently utilized but are severely limited by competing needs to both induce sufficient muscle injury which results in an appropriately severe degree of left ventricular dysfunction yet to also minimize early mortality. MethodsTwenty-eight healthy adult dogs (30{+/-}4 kg, 15/28 male) underwent thoracotomy procedures followed by one of three types of left anterior descending (LAD) coronary artery ligations: Simple LAD (proximal and distal LAD ligation), LAD and Lateral (proximal and distal LAD, distal first diagonal, and proximal first obtuse marginal), and Total LAD Devascularization (TLD including the proximal and distal LAD, proximal aspects of the first, second, and third diagonals, and proximal aspects of first and second epicardial right ventricular branches of the LAD). Invasive and non-invasive hemodynamics were followed until attainment of chronic severe HF, which was achieved when left ventricular ejection fraction (LVEF) was <40% and N-terminal prohormone B-type Natriuretic peptide (NT-proBNP) >900 pmol/L, both for two consecutive measurements. ResultsOverall post-ligation early mortality (within 48 h of coronary ligation) was 18% (n=5/28), and the overall survival rate at 5 weeks post-ligation was 68% (n=19/28). At post-ligation week 1, LVEF was acutely reduced in all animals, from 54{+/-}4 to 38{+/-}1%, p<0.001. Beyond week 1, systolic function recovered over 3-4 weeks and then slowly declined as the left ventricle dilated and both systolic and diastolic function worsened. The magnitude of the initial decline in LVEF predicted ultimate HF development. Dogs with Simple LAD and LAD plus Lateral had excellent and limited survival at 48 hours (100% and 47%, respectively), yet the second group more likely developed chronic HF at 9 - 14 weeks (50% versus 57%, respectively). Dogs with TLD had both improved survival at 48 hours (90%) and excellent development of chronic HF (100%). ConclusionsBy focusing on the LAD only and limiting collateral flow to it by ligating all side branches, optimal survival and attainment of chronic ischemic HF can be achieved in the canine heart.

developmental biology↗

bioRxiv2021

Lager yeast design through meiotic segregation of a fertile Saccharomyces cerevisiae x Saccharomyces eubayanus hybrid

Yeasts in the lager brewing group are closely related and consequently do not exhibit significant genetic variability. Here, an artificial Saccharomyces cerevisiae x Saccharomyces eubayanus tetraploid interspecies hybrid was created by rare mating, and its ability to sporulate and produce viable gametes was exploited to generate phenotypic diversity. Four spore clones obtained from a single ascus were isolated, and their brewing-relevant phenotypes were assessed. These F1 spore clones were found to differ with respect to fermentation performance under lager brewing conditions (15 {degrees}C, 15 {degrees}Plato), production of volatile aroma compounds, flocculation potential and temperature tolerance. One spore clone, selected for its rapid fermentation and acetate ester production was sporulated to produce an F2 generation, again comprised of four spore clones from a single ascus. Again, phenotypic diversity was introduced. In two of these F2 clones, the fermentation performance was maintained and acetate ester production was improved relative to the F1 parent and the original hybrid strain. Strains also performed well in comparison to a commercial lager yeast strain. Spore clones varied in ploidy and chromosome copy numbers, and faster wort fermentation was observed in strains with a higher ploidy. An F2 spore clone was also subjected to 10 consecutive wort fermentations, and single cells were isolated from the resulting yeast slurry. These isolates also exhibited variable fermentation performance and chromosome copy numbers, highlighting the instability of polyploid interspecific hybrids. These results demonstrate the value of this natural approach to increase the phenotypic diversity of lager brewing yeast strains. Contribution to the fieldLager beer fermentations have traditionally been carried out with natural S. cerevisiae x S. eubayanus hybrids. These strains possess both the ability to tolerate low temperatures and the ability to utilize efficiently wort sugars. However, being closely related, strains within the group exhibit limited phenotypic variability. Since the recent discovery of wild strains of S. eubayanus, it has been possible to generate lager yeast hybrids artificially, thereby increasing the genetic and phenotypic diversity of lager brewing strains. Here, to demonstrate the potential for further increased diversity, a constructed tetraploid hybrid was sporulated and spore clones derived from a single ascus were evaluated with respect to fermentation performance (sugar utilization, stress tolerance and volatile aroma synthesis). Meiosis introduced variability in a number of key parameters. One fertile spore clone from this F1 generation was sporulated to introduce further diversity and to demonstrate the potential of clone selection in steering phenotypes in a desirable direction. Genome instability of hybrids was observed, but this can be exploited to further increase diversity. This was demonstrated by assessing performance of variants isolated after ten consecutive rounds of fermentation. The approach allows for the introduction of phenotypic diversity without the need for targeted genetic modification.

microbiology↗

bioRxiv2018

An Empirical Demonstration of Unsupervised Machine Learning in Species Delimitation

One major challenge to delimiting species with genetic data is successfully differentiating species divergences from population structure, with some current methods biased towards overestimating species numbers. Many fields of science are now utilizing machine learning (ML) approaches, and in systematics and evolutionary biology, supervised ML algorithms have recently been incorporated to infer species boundaries. However, these methods require the creation of training data with associated labels. Unsupervised ML, on the other hand, uses the inherent structure in data and hence does not require any user-specified training labels, thus providing a more objective approach to species delimitation. In the context of integrative taxonomy, we demonstrate the utility of three unsupervised ML approaches, specifically random forests, variational autoencoders, and t-distributed stochastic neighbor embedding, for species delimitation utilizing a short-range endemic harvestman taxon (Laniatores, Metanonychus). First, we combine mitochondrial data with examination of male genitalic morphology to identify a priori species hypotheses. Then we use single nucleotide polymorphism data derived from sequence capture of ultraconserved elements (UCEs) to test the efficacy of unsupervised ML algorithms in successfully identifying a priori species, comparing results to commonly used genetic approaches. Finally, we use two validation methods to assess a priori species hypotheses using UCE data. We find that unsupervised ML approaches successfully cluster samples according to species level divergences and not to high levels of population structure, while standard model-based validation methods over-split species, in some instances suggesting that all sampled individuals are distinct species. Moreover, unsupervised ML approaches offer the benefits of better data visualization in two-dimensional space and the ability to accommodate various data types. We argue that ML methods may be better suited for species delimitation relative to currently used model-based validation methods, and that species delimitation in a truly integrative framework provides more robust final species hypotheses relative to separating delimitation into distinct \"discovery\" and \"validation\" phases. Unsupervised ML is a powerful analytical approach that can be incorporated into many aspects of systematic biology, including species delimitation. Based on results of our empirical dataset, we make several taxonomic changes including description of a new species.

evolutionary biology↗

bioRxiv2017

Principal Metabolic Flux Mode Analysis

MotivationIn the analysis of metabolism using omics data, two distinct and complementary approaches are frequently used: Principal component analysis (PCA) and Stoichiometric flux analysis. PCA is able to capture the main modes of variability in a set of experiments and does not make many prior assumptions about the data, but does not inherently take into account the flux mode structure of metabolism. Stoichiometric flux analysis methods, such as Flux Balance Analysis (FBA) and Elementary Mode Analysis, on the other hand, produce results that are readily interpretable in terms of metabolic flux modes, however, they are not best suited for exploratory analysis on a large set of samples.\n\nResultsWe propose a new methodology for the analysis of metabolism, called Principal Metabolic Flux Mode Analysis (PMFA), which marries the PCA and Stoichiometric flux analysis approaches in an elegant regularized optimization framework. In short, the method incorporates a variance maximization objective form PCA coupled with a Stoichiometric regularizer, which penalizes projections that are far from any flux modes of the network. For interpretability, we also introduce a sparse variant of PMFA that favours flux modes that contain a small number of reactions. Our experiments demonstrate the versatility and capabilities of our methodology.\n\nAvailabilityMatlab software for PMFA and SPMFA is available in https://github.com/ aalto-ics-kepaco/PMFA.\n\nContactsahely@iitpkd.ac.in, juho.rousu@aalto.fi, Peter.Blomberg@vtt.fi, Sandra.Castillo@vtt.fi\n\nSupplementary informationDetailed results are in Supplementary files. Supplementary data are available at https://github.com/aalto-ics-kepaco/PMFA/blob/master/Results.zip.

bioinformatics↗