Biology subjects

Goncalves, J. P.

Publications and source records attributed to Goncalves, J. P..

8 recordsLinked to original sources

bioRxiv2026

SigDyn: single-cell mutational signature dynamics

MotivationCancer is driven by mutations that confer tumor cells proliferation advantages but also vulnerabilities with therapeutic potential. Underlying tumor progression are mutational processes that imprint signatures of co-occurring mutation types in tumor cell genomes. The discovery of signatures and their etiologies offers insight into disease mechanisms and treatment opportunities, however acquisition of mutations by tumor cells over time leads to intratumor heterogeneity that may challenge treatment efficacy. Understanding which mutational processes contribute to the evolution of each tumor can reveal therapeutic strategies, but it requires single-cell mutational signature analysis, which remains unexplored. ResultsWe present SigDyn, a framework enabling analysis of tumor mutational signature dynamics at single-cell resolution. To do this, SigDyn combines variant detection, tumor phylogeny inference, and signature identification. We applied SigDyn to scRNA-seq of invasive ductal carcinoma (IDC) and laryngeal squamous cell carcinoma (LSCC). Inferred phylogenies recapitulated tumor markers and progression from normal tissue to tumor metastasis. Tumor-specific processes beyond aging, like cellular disruptions or patient treatments, were only exposed at single-cell level reinforcing the importance of granularity. For IDC, prominence of signature SBS26 with tumor evolution pointed to mismatch repair (MMR) deficiency, supported by downregulation of MMR genes. For LSCC, metastasis enrichment with SBS32-active cells and immune markers suggested influence of immunosuppressant treatment azathioprine, linked to increased risk of SCCs. The dynamics uncovered by SigDyn confirm its potential as a tool to investigate mechanisms of tumor evolution and heterogeneity. Availabilitygithub.com/joanagoncalveslab/sigdyn. Contactjoana.goncalves@tudelft.nl. Supplementary Information: included.

bioinformatics↗

bioRxiv2025

Signatures in CRISPR Mutational Spectra Reveal Role and Interplay of Genes in DNA Repair

MotivationUnderstanding double-strand break (DSB) repair and its disruption is key to decipher genomic instability driving diseases such as cancer and reveal therapeutic avenues. Numerous genes have been linked to DSB repair for the first time in recent genome-wide perturbation studies assessing effects on mutational outcomes. However, the functional roles of most such genes remain poorly understood. Evidence from other studies shows that related genes similarly modulate the frequency of specific mutational outcomes following DSB repair, but analysis has largely ignored the multiplicity of gene functions and cross-talk between pathways. Here, we infer functional roles for candidate genes based on knockout mutational spectra by identifying mutational signatures shared with known genes and then mapping them to DSB repair functions. Signatures are identified using non-negative matrix factorization (NMF) to reflect functional multiplicity and cross-talk. ResultsWe generated mutational spectra for mouse embryonic stem (mES) cells at three target sites across CRISPR knockouts of 307 known and 459 candidate DSB repair genes. Analysis using NMF revealed four mutational signatures associated with homology-directed repair (HDR), microhomology-mediated end-joining (MMEJ), and the initiation and ligation components of non-homologous end-joining (NHEJ). Signatures suggested that candidate genes Dbr1 and Hnrnpk could influence MMEJ and Fanconi anaemia (FA), and that loss of core NHEJ components (e.g. MRN complex or Ku proteins) could shift repair preference towards Ku-independent NHEJ. These findings demonstrate the utility of NMF for characterizing the contribution of genes to repair pathways and provide a foundation to discover new gene functions in DSB repair. Availabilitygithub.com/joanagoncalveslab/MuSpectraNMF.

bioinformatics↗

bioRxiv2025

Oncostratifier: Stratifying Oncogene-addicted Cohorts By Drug Response

Withdrawal StatementThe authors have withdrawn this manuscript because of a disagreement between authors regarding the work. Therefore, the authors do not wish this work to be cited as reference for the project. If you have any questions, please contact the corresponding author.

cancer biology↗

bioRxiv2025

X-CRISP: Domain-Adaptable and Interpretable CRISPR Repair Outcome Prediction

MotivationControlling the outcomes of CRISPR editing is crucial for the success of gene therapy. Since donor template-based editing is often inefficient, alternative strategies have emerged that leverage mutagenic end-joining repair instead. Existing machine learning models can accurately predict end-joining repair outcomes, however: generalisability beyond the specific cell line used for training remains a challenge, and interpretability is typically limited by suboptimal feature representation and model architecture. ResultsWe propose X-CRISP, a flexible and interpretable neural network for predicting repair outcome frequencies based on a minimal set of outcome and sequence features, including microhomologies (MH). Outperforming prior models on detailed and aggregate outcome predictions, X-CRISP prioritised MH location over MH sequence properties such as GC content for deletion outcomes. Through transfer learning, we adapted X-CRISP pre-trained on wild-type mESC data to target human cell lines K562, HAP1, U2OS, and mESC lines with altered DNA repair function. Adapted X-CRISP models improved over direct training on target data from as few as 50 samples, suggesting that this strategy could be leveraged to build models for new domains using a fraction of the data required to train models from scratch. AvailabilityAn implementation of X-CRISP is available at github.com/joanagoncalveslab/xcrisp.

bioinformatics↗

bioRxiv2025

MUSICiAn: Genome-wide Identification of Genes Involved in DNA Repair via Control-Free Mutational Spectra Analysis

MotivationUnderstanding the factors involved in DNA double-strand break (DSB) repair is crucial for the development of targeted anti-cancer therapies, yet the roles of many genes remain unclear. Recent studies show that perturbations of certain genes can alter the distribution of sequence-specific mutations left behind after DSB repair. This suggests that genome-wide screening could reveal novel DSB repair factors by identifying genes whose perturbation causes the mutational distribution spectra observed at a given DSB site to deviate significantly from the wild-type. However, designing proper controls for a genome-wide perturbation screen could be challenging. We explore the idea that a genome-wide screen might allow us to forgo the use of traditional non-targeting controls by reframing the analysis as an outlier detection problem, assuming that most genes have minimal influence on DSB repair. ResultsWe propose MUSICiAn (Mutational Signature Catalogue Analysis), a compositional data analysis method that ranks gene perturbation-specific mutational spectra without controls by measuring deviations from the central tendency in the distributions of all spectra. We show that MUSICiAn can effectively estimate pseudo-controls for the existing Repair-seq dataset, screening 476 genes and 60 non-targeting controls. We further apply MUSICiAn to a genome-wide dataset profiling mutational outcomes induced by CRISPR-Cas9 at three target sites across cells with individual perturbations of 18,406 genes. MUSICiAn successfully recovers known genes, highlights the spliceosome as a lesser-appreciated player in DSB repair, and reveals candidates for further investigation. Availabilitygithub.com/joanagoncalveslab/MUSICiAn.

bioinformatics↗

bioRxiv2024

SNMF: Integrated Learning of Mutational Signatures and Prediction of DNA Repair Deficiencies

MotivationMany tumours show deficiencies in DNA damage response (DDR), which influence tumorigenesis and progression, but also expose vulnerabilities with therapeutic potential. Assessing which patients might benefit from DDR-targeting therapy requires knowledge of tumour DDR deficiency status, with mutational signatures reportedly better predictors than loss of function mutations in select genes. However, signatures are identified independently using unsupervised learning, and therefore not optimised to distinguish between different pathway or gene deficiencies. ResultsWe propose SNMF, a supervised non-negative matrix factorisation that jointly optimises the learning of signatures: (1) shared across samples, and (2) predictive of DDR deficiency. We applied SNMF to mutation profiles of human induced pluripotent cell lines carrying gene knockouts linked to three DDR pathways. The SNMF model achieved high accuracy (0.971) and learned more complete signatures of the DDR status of a sample, further discerning distinct mechanisms within a pathway. Cell line SNMF signatures recapitulated tumour-derived COSMIC signatures and predicted DDR pathway deficiency of TCGA tumours with high recall, suggesting that SNMF-like models can leverage libraries of induced DDR deficiencies to decipher intricate DDR signatures underlying patient tumours. Availabilityhttps://github.com/joanagoncalveslab/SNMF.

bioinformatics↗

bioRxiv2022

ELISL: Early-Late Integrated Synthetic Lethality Prediction in Cancer

Anti-cancer therapies based on synthetic lethality (SL) exploit tumor vulnerabilities for treatment with reduced side effects. Since simultaneous loss-of-function of SL genes causes cell death, tumors with known gene disruptions can be treated by targeting SL partners. Computational selection of promising SL candidates amongst all gene combinations is key to expedite experimental screening. However, current SL prediction models: (i) only use tissue type-specific molecular data, which can be scarce/noisy, limiting performance for some cancers; and (ii) often rely on shared SL patterns across genes, showing sensitivity to prevalent gene selection bias. We propose ELISL, Early-Late Integrated models for SL prediction using forest ensembles. ELISL models ignore shared SL patterns, and integrate context-specific data from cancer cell lines or tumor tissue with context-free functional associations derived from protein sequence. ELISL outperformed existing methods and was more robust to selection bias in 8 cancer types, with prominent contribution from sequence. We found better survival for patients whose tumors carried simultaneous mutations in a BRCA gene together with an ELISL-predicted SL gene from the HH, FGF, or WNT families. ELISL thus arises as a promising strategy to discover SL interactions with therapeutic potential.

bioinformatics↗

bioRxiv2020

Innate fear responses are reflected in the blood epigenome of rhesus macaques.

AO_SCPLOWBSTRACTC_SCPLOWFear and anxiety are complex physiological states aimed at promoting adaptive behaviors. They are also core symptoms of many neuropsychiatric disorders; yet, our knowledge of the underlying biological correlates remains fragmented. Non-human primate models are critical for our understanding of mechanisms associated with complex higher-order behavioral phenotypes. Here we investigated individual variations in innate fear responses to a snake stimulus in free-ranging rhesus macaques and discovered an unusual bimodal distribution of fearful and fearless behavior, likely as a result of an environmental insult by a hurricane. In a translational approach, we discovered a DNA methylation profile associated with fear behavior in these monkeys. We also found evidence that this epigenetic signature is associated with innate fear responses in humans in the form of acoustic startle. Our data highlight the importance and translational utility of non-human primate models for neuropsychiatric research and provide a potential epigenetic signature of innate fear.

neuroscience↗