MTMargret Thorsteinsdottir

Margret Thorsteinsdottir

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Margret Thorsteinsdottir has a total of 184 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Diagnostic plasma protein signatures in sporadic and BRCA1/2 breast cancer patients via targeted proteomics

    by Kristrun Y. Holm, Yassene Mohammed, Valdis Gunnarsdottir Thormar, Finnur F. Eiríksson, Christoph H. Borchers, Stefan Sigurdsson, Jon G. Jonasson, Margret Thorsteinsdottir, Sigridur K. Bodvarsdottir

    Early breast cancer diagnosis is critical, particularly for individuals at hereditary risk. The aim of this study was to identify blood-derived protein bio

  • Open Access

    Outcomes and future activities of the ‘Pan-European network in Lipidomics and EpiLipidomics – EpiLipidNET’

    by Julia Kuligowski, Justine Bertrand-Michel, Amaury Cazenave-Gassiot, Laura Goracci, Pedro Domingues, Matej Orešič, Margret Thorsteinsdottir, Florian Gruber, Egon L. Willighagen, Corinne M. Spickett, et al.

    Background: Lipidomics and its branch epilipidomics are rapidly advancing fields that explore the roles of native and modified lipids (e.g., oxidized, nitrated and halogenated lipid species), respectively, in biological systems. Dysregulation of lipid metabolism and signaling contributes to numerous diseases, including cardiovascular, metabolic, neurodegenerative, and inflammatory conditions. However, multiple challenges, including lack of standardization, limited data integration, and poor clinical translation, hinder progress. To address these, the COST Action EpiLipidNET (CA19105) established a pan-European network fostering collaboration across disciplines to accelerate lipid science and its application to health and disease. Aim of review: This review outlines the achievements of EpiLipidNET over its four-year duration, highlights key scientific contributions across five thematic working groups, and presents the future direction of its ongoing activities. The aim is to demonstrate how a collaborative, interdisciplinary framework can catalyze innovation in lipidomics and epilipidomics, enhance methodological harmonization, support early-career researchers, and bridge the gap between basic science, clinical translation, industry, stakeholders, and public engagement. Key scientific concepts of review: EpiLipidNET structured its networking activities around (i) harmonization of analytical workflows, (ii) development of epilipidomics tools and data integration strategies, (iii) translational studies for clinical lipid biomarkers, (iv) investigation of lipid signaling mechanisms, and (v) dissemination and outreach. The network supported over 460 members globally, launched multiple training schools and scientific missions, produced 110+ publications, and fostered new initiatives in endothelial membrane lipidomics, food lipidomics, plant and algae lipids, and redox lipid biology. Its integrative approach sets a foundation for continued progress toward precision medicine and sustainable health interventions through lipid science.


  • Comparison of the effect of allopurinol and febuxostat on 2,8-dihydroxyadenine in plasma and urine: a clinical trial

    by Hrafnhildur Linnet Runolfsdottir, Unnur Arna Thorsteinsdottir, Steinunn Johannesdottir, Thorunn Oskarsdottir, Inger Maria Schweitz Agustsdottir, Margret Thorsteinsdottir, Runolfur Palsson, Vidar Orn Edvardsson

    Background Adenine phosphoribosyltransferase (APRT) deficiency is a rare, inherited metabolic disorder characterized by abundant urinary excretion of 2,8-dihydroxyadenine (DHA), causing urinary stones and chronic kidney disease. The aim of this study was to examine the effect of allopurinol and febuxostat on plasma levels and urinary excretion of DHA in individuals with APRT deficiency. Methods Adult individuals enrolled in the Icelandic APRT Deficiency Registry were invited to participate in a single-center, open-label, crossover, randomized clinical trial comparing the effect of allopurinol 400 mg/day and 800 mg/day and febuxostat 40 mg/day and 80 mg/day on plasma concentration and urinary excretion of DHA. Results Of 12 participants who initiated the study, 7 (3 females) completed the trial; median (range) age 57.7 (37.3–65.1) years. Off pharmacotherapy, the median plasma DHA was 300 (178–1315) ng/mL. In individuals taking allopurinol 400 mg/day and 800 mg/day, the median plasma DHA was 25 (below the limit of detection [LOD]-95) ng/mL and below the limit of detection (< LOD-92) ng/mL, respectively. On febuxostat 40 mg/day, the median plasma DHA was below the limit of detection (< LOD-35) ng/mL and on 80 mg/day DHA was below the limit of detection in all samples tested. The median urine DHA-to-creatinine ratio was 8.18 (6.21–18.69) mg/mmol off pharmacotherapy and 1.90 (< LOD-4.52) mg/mmol and 0.35 (< LOD-4.32) mg/mmol on allopurinol 400 mg/day and 800 mg/day, respectively. During treatment with febuxostat 40 mg/day and 80 mg/day, the urine DHA-to-creatinine ratio was 0.54 (< LOD-1.33) mg/mmol and below the limit of detection (< LOD-0.64) mg/mmol, respectively. Conclusions The plasma concentration and urinary excretion of DHA decreased markedly on treatment with both study drugs, although febuxostat was more efficacious than allopurinol in both prescribed doses. Trial registration number and date of registration. EudraCT No. 2021-002185–40; https://www.clinicaltrialsregister.eu/ctr-search/search?query=Research+Registry Date on which this record was first entered in the EudraCT database: 2019–03-19. Graphical Abstract


  • Correlation of Plasma and Urine 2,8‐Dihydroxyadenine and Adenine and Clinical Characteristics in Individuals With Adenine Phosphoribosyltransferase Deficiency

    by Unnur A. Thorsteinsdottir, Hrafnhildur L. Runolfsdottir, Vidar O. Edvardsson, Margret Thorsteinsdottir, Runolfur Palsson

    Adenine phosphoribosyltransferase (APRT) deficiency is a rare, autosomal recessive disorder characterized by urinary excretion of the poorly soluble 2,8‐dihydroxyadenine (DHA), leading to kidney stones and chronic kidney disease. Treatment with the xanthine oxidoreductase (XOR) inhibitors allopurinol and febuxostat reduces DHA production. DHA and adenine were measured in 122 paired plasma and urine samples from 26 individuals with confirmed APRT deficiency, using ultra‐performance liquid chromatography–tandem mass spectrometry assays. The relationship between plasma DHA and adenine concentrations, urine DHA‐to‐creatinine (DHA/Cr) and adenine‐to‐creatinine (adenine/Cr) ratios, and age and estimated glomerular filtration rate (eGFR) was evaluated in a subset of 87 paired plasma and urine samples from 23 individuals using Spearman's rank correlation. Allopurinol and febuxostat treatment reduced plasma DHA, with the median (range) concentration decreasing from 249 (123–1315) ng/mL in untreated individuals to below the limit of detection in those receiving higher doses. Plasma adenine increased during XOR inhibitor treatment. In untreated individuals, a strong negative correlation was observed between plasma DHA and eGFR (rs = −0.74, p < 0.0001). Plasma DHA correlated with urine DHA/Cr ratio in individuals treated with allopurinol or febuxostat (rs = 0.65, p < 0.0001), while no significant correlation was observed in untreated individuals (rs = −0.26, p = 0.14). Treatment with XOR inhibitors effectively reduces the plasma concentration and urinary excretion of DHA. The strong correlation between plasma DHA and eGFR, combined with the lack of correlation between plasma DHA and urine DHA/Cr ratio in untreated individuals, suggests that plasma DHA may be a more reliable marker of systemic DHA burden.


  • Design of experiments for the optimization of sample preparation for bottom-up targeted protein LC-MS/MS workflows

    by Szabolcs Szarka, Margret Thorsteinsdottir, Robert Wheller

    Design of experiments (DOE) is a versatile and efficient approach to tackle complex scientific problems. We aimed to assess its feasibility in the optimization of the multistep, involved bottom-up sample preparation for the UPLC-MS/MS analysis of proteins. The model analyte was a human IgG1 monoclonal antibody which was spiked into rat plasma and processed further by reduction, alkylation, and digestion for the subsequent UPLC-MS/MS analysis. The Modde Go software was used for the generation of experimental designs and for processing, analyzing and the interpretation of the data. DOE screening revealed that urea made the biggest improvement on the surrogate peptide responses, while guanidine significantly suppressed them. DOE optimization resulted in a 2-, 10-, 10- and 50-fold response increase for the respective DTLM, FNWY, TPEV and VVSV surrogate peptide even after a short, <3-h sample preparation, when compared to a legacy method that required 2-day preparation. The DOE approach was applied successfully for the comprehensive optimization of eight denaturation, reduction and digestion parameters. DOE was found to be an efficient tool for protein sample preparation optimization, and the predictive power of the DOE models was successfully demonstrated.


  • Open Access

    Optimization of bottom-up proteomic sample preparation by design of experiments for absolute quantification of 257 proteins in human plasma using UPLC-MRM-MS

    by Kari Arnarson, Kristrun Y. Holm, Valdis Gunnarsdottir Thormar, Finnur F. Eiríksson, Yassene Mohammed, Christoph H. Borchers, Sigridur K. Bodvarsdottir, Margret Thorsteinsdottir

    Targeted protein absolute quantification using mass spectrometry holds promise for identifying biomarkers for diagnosis, prognosis, and personalized medicine. However, complex and time-consuming workflows, particularly during sample preparation, present significant bottlenecks. Addressing these challenges is critical for the applicability of absolute quantification of proteins in clinical research settings. We explore optimization strategies for protein digestion in bottom-up proteomics sample preparation. Design of experiments (DoE), a statistical approach for systematically evaluating multiple experimental factors, was used for simultaneous optimization of digestion time, temperature, enzyme-to- protein substrate ratio, and denaturing agent. Furthermore, the lower limit of quantification (LLOQ) for our platform was improved by using the Waters Xevo TQ-XS UPLC-MRM-MS. The integration of automated sample preparation into the workflow enabled reproducible absolute quantification of 257 proteins in human plasma. We successfully reduced protein digestion time from 18 hours (overnight) to 4 hours while maintaining relative digestion efficiency. We improved the sensitivity of the assay via the optimized workflow and were able to quantify proteins that previously fell below the LLOQ. These advancements, combined with automation, provide a practical, efficient, and reproducible workflow suitable for clinical research.


  • Open Access

    Optimization and validation of a UPLC-MS/MS assay for simultaneous quantification of 2,8-dihydroxyadenine, adenine, allopurinol, oxypurinol and febuxostat in human plasma

    by Unnur A. Thorsteinsdottir, Hrafnhildur L. Runolfsdottir, Finnur F. Eiriksson, Inger M. Sch. Agustsdottir, Vidar O. Edvardsson, Runolfur Palsson, Margret Thorsteinsdottir


  • Open Access

    Metabolite Profile Characterization of Cyanobacterial Strains with Bioactivity on Lipid Metabolism Using In Vivo and In Vitro Approaches

    by Tiago Ribeiro, Tiago Ribeiro, Tiago Ribeiro, Kristín Jónsdóttir, Kristín Jónsdóttir, Kristín Jónsdóttir, Rene Hernandez-Bautista, Rene Hernandez-Bautista, Rene Hernandez-Bautista, Natália Gonçalves Silva, et al.

    Cyanobacteria have demonstrated their therapeutic potential for many human diseases. In this work, cyanobacterial extracts were screened for lipid reducing activity in zebrafish larvae and in fatty-acid-overloaded human hepatocytes, as well as for glucose uptake in human hepatocytes and ucp1 mRNA induction in murine brown adipocytes. A total of 39 cyanobacteria strains were grown and their biomass fractionated, resulting in 117 chemical fractions. Reduction of neutral lipids in zebrafish larvae was observed for 12 fractions and in the human hepatocyte steatosis cell model for five fractions. The induction of ucp1 expression in murine brown adipocytes was observed in six fractions, resulting in a total of 23 bioactive non-toxic fractions. All extracts were analyzed by untargeted UPLC-Q-TOF-MS mass spectrometry followed by multivariate statistical analysis to prioritize bioactive strains. The metabolite profiling led to the identification of two markers with lipid reducing activity in zebrafish larvae. Putative compound identification using mass spectrometry databases identified them as phosphatidic acid and aromatic polyketides derivatives—two compound classes, which were previously associated with effects on metabolic disorders. In summary, we have identified cyanobacterial strains with promising lipid reducing activity, whose bioactive compounds needs to be identified in the future.


  • Open Access

    Automated Cancer Diagnostics via Analysis of Optical and Chemical Images by Deep and Shallow Learning

    by Olof Gerdur Isberg, Valentina Giunchiglia, James S. McKenzie, Zoltan Takats, Jon Gunnlaugur Jonasson, Sigridur Klara Bodvarsdottir, Margret Thorsteinsdottir, Yuchen Xiang

    Optical microscopy has long been the gold standard to analyse tissue samples for the diagnostics of various diseases, such as cancer. The current diagnostic workflow is time-consuming and labour-intensive, and manual annotation by a qualified pathologist is needed. With the ever-increasing number of tissue blocks and the complexity of molecular diagnostics, new approaches have been developed as complimentary or alternative solutions for the current workflow, such as digital pathology and mass spectrometry imaging (MSI). This study compares the performance of a digital pathology workflow using deep learning for tissue recognition and an MSI approach utilising shallow learning to annotate formalin-fixed and paraffin-embedded (FFPE) breast cancer tissue microarrays (TMAs). Results show that both deep learning algorithms based on conventional optical images and MSI-based shallow learning can provide automated diagnostics with F1-scores higher than 90%, with the latter intrinsically built on biochemical information that can be used for further analysis.