Biology subjects

Mastromatteo, S.

Publications and source records attributed to Mastromatteo, S..

8 recordsLinked to original sources

bioRxiv2026

Beyond statistical significance: ranking transcription factor binding motifs by effect size

Chromatin immunoprecipitation-sequencing (ChIP-seq) has wide use in identifying transcription factor binding sites. DNA sequence motifs specific to a targeted transcription factor occur more frequently near ChIP-seq peak centres. The most common approach to quantifying relative motif enrichment ranks motifs by p-value. Because sample sizes can vary substantially across examined motifs, p-value magnitudes may reflect this heterogeneity rather than the biological effect of interest. As alternatives, we considered four ranking methods based on effect sizes: (a) a modified Cliffs delta, (b) the lower bound of a frequentist asymptotic confidence interval, (c) the lower bound of a frequentist finite-sample confidence interval, and (d) the lower bound of a Bayesian credible region. Through extensive simulations, the four alternatives better recovered the simulated central-enrichment ordering under heterogeneous sample sizes. Using published ChIP-seq data for GATA3, the effect size methods ranked the known targeted motif highest, even compared to highly similar motifs for other GATA family members, while p-value ranking did not. In a separate SRF application, all four alternative methods also consistently ranked the known motif highest. We recommend the asymptotic confidence interval lower bound for its simplicity, ease of implementation, and intuitive interpretation. The software is freely available (https://github.com/ScottMastro/motif-ranking).

bioinformatics↗

bioRxiv2025

Beyond reference bias: Making pangenomes accessible with PangyPlot

Linear reference genomes have standardized genomics research but remain limited by reference bias, which skews read mapping and variant discovery. This bias can distort the interpretation of genetic variation, particularly for populations that are genetically distant from the reference. Pangenome graphs, such as those generated by the Human Pangenome Reference Consortium (HPRC), mitigate this limitation by integrating diverse haplotypes into a unified representation of human genetic variation. However, the complexity of graph-based data and the lack of intuitive visualization tools have hindered broader adoption. Here we introduce PangyPlot, a genome browser that simplifies exploration of pangenome graphs by retaining linear-style navigation, integrating gene annotations, abstracting complex variation into interpretable structures, and employing a dynamic, physics-based layout optimization engine. We demonstrate its utility by constructing a chromosome 7 graph from 101 individuals with cystic fibrosis (CF), capturing a broad spectrum of genetic variation. Using PangyPlot, we visualized CF-relevant loci and compared results with existing graph viewers, highlighting its ability to display both base-level and large structural variation. With an additional 64 PacBio HiFi assemblies, we fine-mapped a repeat-dense CF modifier locus on chromosome 5, where PangyPlot was used in conjunction with graph-based analysis to identify a repeat expansion in the 5' end of EXOC3 that may promote G-quadruplex formation and affect gene expression. Together, these examples demonstrate PangyPlot s capacity to make populationlevel variation interpretable. To support broader use of graph-based resources, we also released a live public instance of PangyPlot preloaded with HPRC data (https://pangyplot.research.sickkids.ca/).

genomics↗

bioRxiv2025

GWAS SVatalog: a visualization tool to aid fine-mapping of GWAS loci with structural variations

BackgroundGenome-wide association studies (GWAS) have been successful in identifying single nucleotide polymorphisms (SNPs) associated with phenotypic traits. However, SNPs form an incomplete set of variation across the genome and since a large percentage of GWAS-significant SNPs lie in non-coding regions, their impact on a given trait is difficult to decipher. Recognizing whether these SNPs are tagging other polymorphisms, like structural variations (SV), is an important step towards understanding the putative causal variation at GWAS loci. ResultsHere, we develop GWAS SVatalog (https://svatalog.research.sickkids.ca/), a novel open-source web tool that computes and visualizes linkage disequilibrium (LD) between SVs and GWAS-associated SNPs throughout the human genome. The tool combines GWAS Catalogs SNP-trait association data across 14,479 phenotypes with LD statistics calculated between 35,732 SVs and 116,870 SNPs identified in 101 whole-genome long-read sequences. We use GWAS SVatalog to identify SVs that may explain GWAS loci for iron levels, refractive error, and Alzheimers disease, where previously SNPs were unable to provide a causal explanation. ConclusionsGWAS SVatalog advances the fine-mapping of GWAS loci with structural variations, enabling researchers to associate 35,732 common SVs with 14,479 phenotypes, accelerating the understanding of disease etiology.

bioinformatics↗

bioRxiv2022

High Quality Phasing Using Linked-Read Whole Genome Sequencing of Patient Cohorts Informs Genetic Understanding of Complex Traits

Phasing of heterozygous alleles is critical for interpretation of cis-effects of disease-relevant variation. For population studies, phase is often inferred from external data but read-based phasing approaches that span long genomic distances would be more accurate because they enable both genotype and phase to be obtained from a single dataset. To demonstrate how read-based phasing can provide functional insights, we sequenced 477 individuals with Cystic Fibrosis (CF) using linked-read sequencing. We benchmark read-based phasing with different short- and long-read sequencing technologies, prioritize linked-read technology as the most informative and produce a benchmark phase call set from reference sample HG002 for the community. The 477 samples display an average phase block N50 of 4.39 Mb. We use these samples to construct a graph representation of CFTR haplotypes, which facilitates understanding of complex CF alleles. Fine-mapping and phasing of the chr7q35 trypsinogen locus associated with CF meconium ileus demonstrates a 20 kb deletion and a PRSS2 missense variant p.Thr8Ile (rs62473563) independently contribute to meconium ileus risk (p=0.0028, p=0.011, respectively) and are PRSS2 pancreas eQTLs (p=9.5e-7 and p=1.4e-4, respectively), explaining the mechanism by which these polymorphisms contribute to CF. Phase enables access to haplotypes that can be used for genome graph or reference panel construction, identification of cis-effects, and for understanding disease associated loci. The phase information from linked-reads provides a causal explanation for variation at a CF-relevant locus which also has implications for the genetic basis of non-CF pancreatitis to which this locus has been reported to contribute.

genetics↗

bioRxiv2020

LocusFocus: A web-based colocalization tool for the annotation and functional follow-up of GWAS

Genome-wide association studies (GWAS) have primarily identified trait-associated loci in the non-coding genome. Colocalization analyses of SNP-level associations from GWAS with expression quantitative trait loci (eQTL) evidence enable the generation of hypotheses about responsible mechanism, genes and tissues of origin to guide functional characterization. Here, we present a web-based colocalization browsing and testing tool named LocusFocus (https://locusfocus.research.sickkids.ca). LocusFocus formally tests colocalization using our established Simple Sum method to identify the most relevant genes and tissues for a particular GWAS locus in the presence of high linkage disequilibrium and/or allelic heterogeneity. Full documentation and source code for LocusFocus are publicly available.

genomics↗

bioRxiv2019

A competence-regulated toxin-antitoxin system in Haemophilus influenzae

Natural competence allows bacteria to respond to environmental and nutritional cues by taking up free DNA from their surroundings, thus gaining both nutrients and genetic information. In the Gram-negative bacterium Haemophilus influenzae, the genes needed for DNA uptake are induced by the CRP and Sxy transcription factors in response to lack of preferred carbon sources and nucleotide precursors. Here we show that one of these genes, HI0659, encodes the antitoxin of a competence-regulated toxin-antitoxin operon ( toxTA), likely acquired by horizontal gene transfer from a Streptococcus species. Deletion of the putative toxin (HI0660) restores uptake to the antitoxin mutant. The full toxTA operon was present in only 17 of the 181 strains we examined; complete deletion was seen in 22 strains and deletions removing parts of the toxin gene in 142 others. In addition to the expected Sxy-and CRP-dependent-competence promoter, HI0659/660 transcript analysis using RNA-seq identified an internal antitoxin-repressed promoter whose transcription starts within toxT and will yield nonfunctional protein. We propose that the most likely effect of unopposed toxin expression is non-specific cleavage of mRNAs and arrest or death of competent cells in the culture. Although the high frequency of toxT and toxTA deletions suggests that this competence-regulated toxin-antitoxin system may be mildly deleterious, it could also facilitate downregulation of protein synthesis and recycling of nucleotides under starvation conditions. Although our analyses were focused on the effects of toxTA, the RNA-seq dataset will be a useful resource for further investigations into competence regulation.\n\nABBREVIATED SUMMARYThe competence regulon of Haemophilus influenzae includes an unprecedented toxin/antitoxin gene pair. When not opposed by antitoxin, the toxin completely prevents DNA uptake but causes only very minor decreases in cell growth and competence gene expression. The TA gene pair was acquired by horizontal gene transfer, and the toxin gene has undergone repeated deletions in other strains.

microbiology↗

bioRxiv2018

VIKNGS: A C++ VARIANT INTEGRATION KIT FOR NEXTGENERATION SEQUENCING ASSOCIATION ANALYSIS

MotivationIntegration of next generation sequencing data (NGS) across different research studies can improve the power of genetic association testing by increasing sample size and can obviate the need for sequencing controls. Unfortunately, if differential genotype uncertainty across studies is not accounted for, combining data sets can also produce spurious association results. The robust variance score statistic (RVS) for genetic association of rare and common variants has been shown to effectively adjust for bias caused by the differences in read depth in case-control genetic association studies when the two groups were sequenced using different experimental designs. To enable consortium research, the aggregation of several data sets for genetic association analysis of quantitative and binary traits with covariate adjustment is required, and we developed the Variant Integration Kit for NGS (VikNGS) that expands the functionality of RVS (vRVS) for this purpose.\n\nResultsVikNGS is a fast and computationally efficient cross-platform software package that provides an implementation for vRVS, as well as conventional rare and common variant genotype-based association analysis approaches. The package includes a graphical user interface that contains power simulation functionality and data visualization tools.\n\nAvailability and ImplementationThe VikNGS package can be downloaded at http://www.tcag.ca/tools/index.html\n\nDocumentation can be found at https://VikNGSdocs.readthedocs.io/en/latest/\n\nContactlisa.strug@sickkids.ca\n\nSupplementary informationSupplementary data are available at Bioinformatics online.

genetics↗

bioRxiv2018

A competence-regulated toxin-antitoxin system in Haemophilus influenzae

Natural competence allows bacteria to respond to environmental and nutritional cues by taking up free DNA from their surroundings, thus gaining nutrients and genetic information. In the Gram-negative bacterium Haemophilus influenae, the DNA uptake machinery is induced by the CRP and Sxy transcription factors in response to lack of preferred carbon sources and nucleotide precursors. Here we show that HI0659--which is absolutely required for DNA uptake-- encodes the antitoxin of a competence-regulated toxin-antitoxin operon ( toxTA), likely acquired by horizontal gene transfer from a Streptococcus species. Deletion of the toxin restores uptake to the antitoxin mutant. In addition to the expected Sxy-and CRP-dependent-competence promoter, transcript analysis using RNA-seq identified an internal antitoxin-repressed promoter whose transcription starts within toxT and will yield nonfunctional protein. We present evidence that the most likely effect of unopposed toxin expression is non-specific cleavage of mRNAs and arrest or death of competent cells in the culture, and we show that the toxin gene has been inactivated by deletion in many H. influenzae strains. We suggest that this competence-regulated toxin-antitoxin system may facilitate downregulation of protein synthesis and recycling of nucleotides under starvation conditions, or alternatively be a simple genetic parasite.

microbiology↗