Biology subjects

Sammut, S.-J.

Publications and source records attributed to Sammut, S.-J..

9 recordsLinked to original sources

bioRxiv2026

DNA methylation and hydroxymethylation quantification using vibrational spectroscopy

Global quantification of DNA cytosine modifications, including 5-methylcytosine (5-mC) and 5-hydroxymethylcytosine (5-hmC), is important for understanding cancer biology, though established methods require multi-step workflows and costly instrumentation. Here we show that attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectroscopy combined with regression modelling enables rapid, label-free, and non-destructive quantification of both modifications from DNA samples. Using Adenomatous Polyposis Coli (APC) promoter DNA standards spanning 0-100% modification, we identified modification-sensitive spectral features and observed that 5-hmC produces greater spectral changes than 5-mC. A univariate peak-ratio approach yielded strong linearity for both modifications (R2 = 0.97), while partial least squares regression (PLSR) improved quantification accuracy to R2 = 0.99 (RMSE = 2.6%) for 5-hmC and R2 = 0.97 (RMSE = 5.7%) for 5-mC. In composite mixtures containing all three cytosine states, 5-hmC remained highly quantifiable (R2 = 0.97; RMSE = 5.1%), while 5-mC accuracy decreased (R2 = 0.90; RMSE = 9.6%), consistent with the greater spectral distinctiveness secondary to the hydroxymethyl group. Transferability was assessed using circulating tumour DNA (ctDNA), short cell-free DNA fragments shed from tumour cells into the bloodstream, comprising multiplexed reference material spanning seven genomic regions and a polydisperse fragment-length distribution (155-220 bp). After domain adaptation between synthetic and ctDNA spectra, we obtained a quantitative methylation calibration with R2 = 0.98 and RMSE = 5.2% under cross-validation. These results support ATR-FTIR spectroscopy as a viable platform for global cytosine modification quantification and establish proof-of-concept applicability to ctDNA analysis.

biochemistry↗

bioRxiv2026

Tumour marker analysis using a machine learning assisted vibrational spectroscopy approach

Tumour biomarkers such as CA125, CA15-3, CA19-9, AFP and CEA are routinely used in the oncology clinic to diagnose cancer, monitor response to therapy, and detect relapse. However, their quantification depends on immunoassay-based methods that are time-consuming, reagent-dependent, and poorly suited to resource-limited settings. Here, we present a machine learning-assisted ATR-FTIR spectroscopy approach for label-free tumour biomarker analysis to enable simple and rapid quantification at the bedside. Using principal component analysis (PCA), we first demonstrate that these five clinically relevant biomarkers are spectrally separable, with the protein-associated region (1200-1700 cm-1) providing the greatest discriminative information. We then develop partial least squares regression (PLSR) models to quantify CA125 in phosphate-buffered saline (R2 = 0.95) and in human serum across a clinically relevant concentration range, achieving reliable predictions at and above the clinical decision threshold of 35 U/mL. A semi-quantitative classification model further demonstrated robust identification of elevated CA125, with a macro-average sensitivity of 0.86 and specificity of 0.92. These results support ATR-FTIR spectroscopy as a rapid, reagent-free platform for cancer biomarker monitoring, with potential utility in resource-limited settings.

biochemistry↗

bioRxiv2026

DNA fragment length analysis using machine learning assisted vibrational spectroscopy

DNA length analysis is essential for genomic workflows including next-generation sequencing and fragmentomics based diagnostics. Conventional approaches typically require large, expensive instrumentation and sample-destructive protocols with long processing times. Here we present a rapid, label-free approach integrating vibrational spectroscopy with deep learning to quantify DNA fragment length distributions. We demonstrate that ATR-FTIR and Raman spectroscopy capture length-dependent spectral features arising from phosphate backbone, nucleobase, and structural vibrations. Machine learning models trained on spectra acquired from purified monodisperse DNA (50-300 bp) predicted DNA length with high accuracy (R2=0.92-0.94), with multimodal fusion improving performance to R2=0.96. A convolutional neural network trained on 35 DNA mixtures comprising molecules of different lengths also successfully deconvoluted their fragment length profile. Transfer learning enabled adaptation to biological samples, achieving low prediction error (RMSE=0.3-7.2%, {Delta}=12 bp). Importantly, the method requires only 4 L sample and 15 minutes passive drying, with no consumables beyond cleaning materials, and allows full sample recovery. This establishes vibrational spectroscopy as a scalable alternative for DNA length quantification.

bioinformatics↗

bioRxiv2025

Interventionally-guided representation learning for robust and interpretable AI models in cancer medicine

Machine learning models hold promise in cancer medicine but often lack robustness and interpretability. We introduce a new class of model for high-dimensional molecular data that incorporate interventional auxiliary information to learn latent representations that are informative and interpretable by design. By using causal signals from genetic loss-of-function screens, our approach generates representations that generalize well across data distributions and biological contexts. In cancer cell line datasets, we show that causal guidance enables "zero-shot" transfer to cancer types unseen during training. Moreover, models trained solely on cell line data translate effectively to clinical cohorts, demonstrating strong "bench-to-bedside" generalization without fine-tuning. This strategy highlights a scalable way to leverage tractable laboratory assays for clinical modeling. More broadly, our results establish how integrating causal biological information within generative frame-works enhances data efficiency, interpretability, and robustness, opening avenues for a new generation of scientifically informed AI models in molecular medicine.

bioinformatics↗

bioRxiv2024

The expression patterns of different cell types and their interactions in the tumor microenvironment are predictive of breast cancer patient response to neoadjuvant chemotherapy

The tumor microenvironment (TME) is a complex ecosystem of diverse cell types whose interactions govern tumor growth and clinical outcome. While multiple studies have extensively charted the TMEs impact on immunotherapy, its role in chemotherapy response remains less explored. To address this, we developed DECODEM (DEcoupling Cell-type-specific Outcomes using DEconvolution and Machine learning), a generic computational framework leveraging cellular deconvolution of bulk transcriptomics to associate gene expression of individual cell types in the TME with clinical response. Employing DECODEM to analyze gene expression of breast cancer patients treated with neoadjuvant chemotherapy across three bulk cohorts, we find that the expression of specific immune cells (myeloid, plasmablasts, B-cells) and stromal cells (endothelial, normal epithelial, CAFs) are highly predictive of chemotherapy response, achieving the same prediction levels as the expression of malignant cells. Notably, ensemble models integrating the estimated expression of different cell types perform best and outperform models built on the original tumor bulk expression. These findings and the models generalizability are further tested and validated in two single-cell (SC) cohorts of triple negative breast cancer. To investigate the possible role of immune cell-cell interactions (CCIs) in mediating chemotherapy response, we extended DECODEM to DECODEMi to identify such key functionally important CCIs, validated in SC data. Our findings highlight the importance of active pre-treatment immune infiltration for chemotherapy success. DECODEM and DECODEMi are made publicly available to facilitate studying the role of the TME in mediating response in a wide range of cancer indications and treatments.

bioinformatics↗

bioRxiv2024

Chemotherapy-driven de novo Wnt pathway activation dictates a dynamic shift to a drug-tolerant state in breast cancer cells

The efficacy of chemotherapy is often hindered by the enrichment of a population of cancer cells that enter a drug-tolerant persister (DTP) state, mimicking embryonic diapause, yet the underlying mechanisms of this transition remain poorly understood. This study demonstrates that both parental and chemotherapy-induced Wnt-active (WntHigh) cells in Triple-negative breast cancer exhibit transcriptional and functional properties characteristic of DTP cells, including a diapause transcriptional signature, reduced MYC expression, reversible restricted proliferation, and pronounced chemoresistance. Our findings reveal that the de novo activation of the Wnt signaling pathway, triggered by the transcriptional upregulation of components essential for canonical Wnt ligand-secretion and -activation, is critical for enriching the diapause-DTP (DTPDiap) population across various chemotherapy regimens. The diapause-DTP/WntHigh population can be selectively ablated by concomitant, rather than sequential, pharmacological inhibition of Wnt ligand-secretion alongside chemotherapy, highlighting new vulnerabilities in DTPDiap cell-emergence and potentially yielding a therapeutic opportunity against DTPs. This study shows that activation of Wnt signaling pathway is sufficient and necessary for the induction of a DTPDiap state and enhances our understanding of the introductory mechanisms driving DTP cell-enrichment upon chemotherapy.

cancer biology↗

bioRxiv2023

A large-scale retrospective study in metastatic breast cancer patients using circulating tumor DNA and machine learning to predict treatment outcome and progression-free survival

PurposeMonitoring levels of circulating tumor-derived DNA (ctDNA) represents a non-invasive snapshot of tumor burden and potentially clonal evolution. Here we describe how a novel statistical model that uses serial ctDNA measurements from shallow whole genome sequencing (sWGS) in metastatic breast cancer patients produces a rapid and inexpensive assessment that is predictive of treatment response and progression-free survival. Patients and MethodsA cohort of 188 metastatic breast cancer patients had DNA extracted from serial plasma samples (total 1098, median=4, mean=5.87). Plasma DNA was assessed using sWGS and the tumor fraction in total cell free DNA estimated using ichorCNA. This approach was compared with ctDNA targeted sequencing and serial CA 15-3 measurements. The longitudinal ichorCNA values were used to develop a Bayesian learning model to predict subsequent treatment response. ResultsWe identified a transition point of 7% estimated tumor fraction to stratify patients into different categories of progression risk using ichorCNA estimates and a time-dependent Cox model, validated across different breast cancer subtypes and treatments, outperforming the alternative methods. We then developed a Bayesian learning model to predict subsequent treatment response with a sensitivity of 0.75 and a specificity of 0.66. ConclusionIn patients with metastatic breast cancer, sWGS of ctDNA and ichorCNA provide prognostic and predictive real-time valuable information on treatment response across subtypes and therapies. A prospective large-scale clinical trial to evaluate clinical benefit of early treatment changes based on ctDNA levels is now warranted.

genomics↗

bioRxiv2018

Detection of residual disease after neoadjuvant therapy in breast cancer using personalized circulating tumor DNA analysis

Accurate detection of minimal residual disease (MRD) can guide individualized management of early stage cancer patients, but current diagnostic approaches lack adequate sensitivity. Circulating tumor DNA (ctDNA) analysis has shown promise for recurrence monitoring but MRD detection immediately after neoadjuvant therapy or surgical resection has remained challenging. We have developed TARgeted DIgital Sequencing (TARDIS) to simultaneously analyze multiple patient-specific cancer mutations in plasma and improve sensitivity for minute quantities of residual tumor DNA. In 77 reference samples at 0.03%-1% mutant allele fraction (AF), we observed 93.5% sensitivity. Using TARDIS, we analyzed ctDNA in 34 samples from 13 patients with stage II/III breast cancer treated with neoadjuvant therapy. Prior to treatment, we detected ctDNA in 12/12 patients at 0.002%-1.04% AF (0.040% median). After completion of neoadjuvant therapy, we detected ctDNA in 7/8 patients with residual disease observed at surgery and in 1/5 patients with pathological complete response (odds ratio, 18.5, Fishers exact p=0.032). These results demonstrate high accuracy for a personalized blood test to detect residual disease after neoadjuvant therapy. With additional clinical validation, TARDIS could identify patients with molecular complete response after neoadjuvant therapy who may be candidates for nonoperative management.\n\nOne Sentence SummaryA personalized ctDNA test achieves high accuracy for residual disease.

genomics↗

bioRxiv2018

Next Generation-Targeted Amplicon Sequencing (NG-TAS): An optimised protocol and computational pipeline for cost-effective profiling of circulating tumour DNA

Circulating tumour DNA (ctDNA) detection and monitoring has enormous potential clinical utility in oncology. We describe here a fast, flexible and cost-effective method to profile multiple genes simultaneously in low input cell-free DNA (cfDNA): Next Generation-Targeted Amplicon Sequencing (NG-TAS). We designed a panel of 377 amplicons spanning 20 cancer genes and tested the NG-TAS pipeline using cell-free DNA from two hapmap lymphoblastoid cell lines. NG-TAS consistently detected mutations in cfDNA when mutation allele fraction was >1%. We applied NG-TAS to a clinical cohort of metastatic breast cancer patients, demonstrating its potential in monitoring the disease. The computational pipeline is available at: https://github.com/cclab-brca/NGTAS_pipeline.

genomics↗