Ai-Powered Multi-Omics Integration for Precision Medicine
Presentation number: PP25
Abstract number: 49-ISABS-2026
TOWARDS CLINICAL INTEGRATION OF CARDIOVASCULAR PHARMACOGENOMICS: THE CARDIOPHARMAGENET APPROACH
Ašić Adna1, Devaux Yvan2, Sopić Miron3, Jusić Amela4, Spahić Lemana1, Marjanović Damir5, Tashkov Konstantin6, Katsila Theodora7, Džuho Amra1
1Verlab Research Institute for Biomedical Engineering, Medical Devices and Artificial Intelligence, Sarajevo, Bosnia and Herzegovina; 2Cardiovascular Research Unit, Department of Precision Health, Luxembourg Institute of Health, Strassen, Luxembourg; 3Department of Medical Biochemistry, Faculty of Pharmacy, University of Belgrade, Belgrade, Serbia; 4Quant Biomarkers, Basel, Switzerland; 5Institute for Anthropological Research, University of Zagreb, Zagreb, Croatia; 6Faculty of Pharmacy, Medical University, Sofia, Bulgaria; 7Institute of Chemical Biology, National Hellenic Research Foundation, Athens, Greece
adna.a@verlabinstitute.com
Cardiovascular diseases remain the leading cause of mortality worldwide, while the integration of pharmacogenomics (PGx) into clinical practice is still fragmented and slow across Europe due to the lack of harmonized guidelines, limited clinical implementation, and inconsistent policy frameworks. Furthermore, the absence of centralized data infrastructures and insufficient use of advanced digital tools, including AI, further hinder the translation of PGx research into routine care. COST Action CA24165 Network for Cardiovascular Pharmacogenomics and Precision Medicine (CardioPharmaGENET) addresses these challenges by establishing a multidisciplinary, pan-European network aimed at harmonizing the generation, analysis, storage, and clinical use of cardiovascular PGx data, while creating a centralized knowledge hub and delivering evidence-based recommendations for policymakers and clinicians. The main objective of the Action is to enable the equitable and consistent implementation of PGx in cardiovascular medicine through coordinated research, capacity building, and stakeholder engagement. The work is structured into five dedicated Working Groups: WG1 Current landscape and status of PGx in cardiovascular medicine; WG2 Pharmacogenetic guideline development and optimization; WG3 AI/ML use in PGx of cardiovascular medicine; WG4 Policy regulation of PGx guidelines and impact analysis across Europe; and WG5 Communication, dissemination and exploitation. By integrating expertise from diverse disciplines and fostering collaboration across countries and sectors, the Action aims to bridge the gap between research and clinical practice, ultimately improving patient outcomes and reducing healthcare inequalities. Within only five months of implementation, CardioPharmaGENET has already engaged over 230 participants from 38 countries, demonstrating strong community uptake and positioning the network as a key driver of future advancements in personalized CVD medicine in Europe.
Keywords: pharmacogenomics, personalized medicine, cardiovascular medicine, artificial intelligence, COST Action
Presentation number: PP26
Abstract number: 28-ISABS-2026
IDENTIFICATION AND CHARACTERIZATION OF RARE DE NOVO CODING MUTATIONS IN CONGENITAL ANOMALIES OF THE KIDNEY AND URINARY TRACT (CAKUT) USING WHOLE EXOME SEQUENCING
Bajt Patricija1,3, Gelemanović Andrea2, Kelam Nela1,3, Racetin Anita1,3, Todorović Petar1,3, Pavlović Nikola1,3, Rakić Tomislav1,3, Arapović Adela4, Simičić Majce Ana4, Vukojević Katarina1,2,3
1Department of Anatomy, Histology and Embryology, University of Split School of Medicine, Split, Croatia; 2Mediterranean Institute for Life Sciences, University of Split, Split, Croatia; 3Center for Translational Research in Biomedicine, University of Split School of Medicine, Split, Croatia; 4Department of Pediatrics, University Hospital of Split, Split, Croatia
patricija.bajt@mefst.hr
Congenital anomalies of the kidney and urinary tract (CAKUT) represent a group of developmental disorders and are one of the leading causes of chronic kidney disease in children. The aim of this study was to identify rare de novo coding mutations associated with CAKUT and highlight their importance in understanding disease mechanisms and improving genetic diagnostics. Samples are obtained from human peripheral blood, DNA is extracted, and Whole Exome Sequencing (WES) is performed to generate genetic data. The study focused on 13 affected probands, comparing their genotypes with those of their unaffected family members to identify de novo mutations. Variants that were common, non-coding, or clinically benign/uncertain were excluded to focus on rare, potentially pathogenic changes. Filtered de novo mutations were then analyzed by mutation type to characterize the genetic alterations in the cohort. Mutation counts per gene and proband were visualized using a CDS-length–normalized heatmap to allow comparison across genes of different sizes. In the cohort of 13 probands, missense mutations were the most frequent type, followed by other functional variants such as frameshift and splice-site mutations. SNVs were significantly more common than indels, and base substitution analysis showed a predominance of specific transitions (e.g., C>T) across the cohort. Using a CDS-length–normalized heatmap, the distribution of mutations in the top 20 genes was visualized for each proband, highlighting which genes were proportionally most affected by de novo mutations. The identified candidate genes represent a valuable resource for future functional studies and may contribute to improved diagnosis and personalized clinical management.
Keywords: CAKUT, whole exome sequencing, de novo variants, SNV, genetic analysis
Presentation number: PP27
Abstract number: 61-ISABS-2026
FROM BIOMARKER DISCOVERY TO REGENERATIVE THERAPY: THE DUAL POTENTIAL OF MENSTRUAL BLOOD-DERIVED MSCS AND EXTRACELLULAR VESICLES
Bernotiene Eiva1,2, Vaiciuleviciute Raminta1,3, Brennan Kieran3, Uzieliene Ilona1, Pachaleva Jolita1, Kugaudaite Gabija1, Bakutyte Ieva1, Lebedis Ignas1, Kasilovskiene Zaneta4, Piesiniene Lina5, McGee Margaret3
1Regenerative Medicine Department, Centre for Innovative Medicine, Vilnius, Lithuania; 2Faculty of Fundamental Sciences, VilniusTech, Vilnius, Lithuania; 3Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Dublin, Ireland; 4Medical Center “MAXMEDA”, Vilnius, Lithuania; 5Nanodiagnostika, Ltd., Vilnius, Lithuania
eiva.bernotiene@imcentras.lt
The identification of novel mesenchymal stromal cell (MSC) sources for biomarker discovery and regenerative application represents a key priority in translational medicine. Menstrual blood is a unique, non-invasive, and ethically acceptable source with utility in both diagnostics and regenerative medicine. Menstrual blood-derived mesenchymal stromal cells (MenSCs) and their extracellular vesicles (EVs) have emerged as promising tools in reproductive medicine, where both applications are urgently needed. Proteomic profiling of menstrual blood serum EVs reveals distinct molecular signatures in unexplained infertility compared to fertile controls, capturing alterations in cell adhesion, immune response, apoptosis, fibrosis, metabolism, and oxidative stress. These profiles enable patient stratification into endotypes, supporting personalised diagnostic strategies. Transcriptomic analysis further confirmed changes in unexplained infertility. From a regenerative perspective, MenSCs exhibit high proliferative capacity, multipotency, and an immunomodulatory secretome. MenSC-derived EVs might hold potential to restore endometrial receptivity and uterine microenvironments central to implantation failure — a key unmet need in infertility management. Beyond reproductive medicine, MenSC paracrine factors may stimulate regenerative responses in different tissues. We demonstrated that MenSC-EVs enhance extracellular matrix production, chondrogenic differentiation, and protect cartilage from degradation, extending their therapeutic reach to musculoskeletal disorders. The translational potential of MenSC-derived products as accessible biomarker sources and cell-free therapeutics positions menstrual blood as a compelling precision medicine resource for reproductive and musculoskeletal conditions alike. Funded by HORIZON-WIDERA-2021-ACCESS-03-01 program 382 project No. 101079489-TWINFLAG and EP Permed call 2024, project PERFERT 2025-2028
Keywords: menstrual blood, extracellular vesicles, proteomic profiling, unexplained infertility, cartilage regeneration
Presentation number: PP28
Abstract number: 109-ISABS-2026
A MACHINE-LEARNING-GUIDED LIPIDOMICS PIPELINE FOR IDENTIFYING CANDIDATE LIPID METABOLIC CHANGES ASSOCIATED WITH ACQUIRED VEMURAFENIB RESISTANCE IN BRAFV600E-MUTANT COLORECTAL CANCER
Bošnjaković Anja1, Šunić Iva1, Karlović Nina1, Meščić Macan Andrijana1, Žepić Ines1, N. Rechberger Gerald2, Züllig Thomas2, Lovrić Mario1, Sedić Mirela1, Bočkor Luka1
1Centre for Applied Bioanthropology, Institute for Anthropological Research, Zagreb; Croatia; 2Institut für Molekulare Biowissenschaften, Karl-Franzens-Universität Graz, Austria
anja.bosnjakovic@inantro.hr
BRAFV600E-mutant metastatic colorectal cancer is a clinically aggressive subgroup with limited benefit from vemurafenib monotherapy because of rapid development of resistance. The aim of this study was to identify lipidomic features linked with acquired resistance to BRAF inhibition and to build an analysis pipeline for selecting candidate lipid metabolic pathways for further studies. Vemurafenib-resistant RKO BRAFV600E colon cancer cells were generated by gradual exposure of parental cells to increasing drug concentrations over 6 months until stable resistance to 11.52 µM was achieved. Resistance was confirmed by MTT assay and morphological evaluation. Lipids were extracted using the Matyash protocol and analyzed by UHPLC-QTOF-MS. Untargeted lipidomics data from parental and resistant cells, under treated and control conditions, were processed and explored using L1/L2 logistic regression and linear SVM models. Internal cross-validation showed good separation between parental and resistant cells and treatment subgroups, indicating group-related biological differences in the lipidomic data. HexCer 34:0;2O, LPI 20:4 and FA 18:1 changed with vemurafenib treatment in both parental and resistant cells, indicating a shared response to drug exposure. In contrast, many PC, PI and LPC species showed different treatment- related changes between parental and resistant cells, suggesting a possible link to acquired resistance. To extend these findings, BioPAN analysis was performed on the same experimental contrasts. Combining BioPAN outputs with lipids identified by our ML analysis highlighted recurrent changes in membrane lipid remodeling and lipid storage pathways, with candidate genes including PLA2, LPCAT, MBOAT7, DGAT2, PEMT and CHPT1. These findings do not identify definitive targets, but they provide a practical framework for generating hypotheses and prioritizing lipid-related pathways for follow-up validation in BRAFV600E-mutant colorectal cancer.
Keywords: BRAFV600E, colorectal cancer, lipidomics, vemurafenib resistance, machine learning
Presentation number: PP29
Abstract number: 80-ISABS-2026
FC-RECEPTOR AFFINITY CHROMATOGRAPHY: A TOOL FOR IVIG THERAPEUTICS PROFILING
Dončević Lucija1,1
1Genos d.o.o., Zagreb, Croatia
ldoncevic@genos.hr
Intravenous immunoglobulin (IVIg) therapeutics are complex biological mixtures of immunoglobulin G (IgG) requiring robust analytical methods to resolve their functional and structural heterogeneity. First established for immunodeficiencies, IVIg concentrates are widely used for immunomodulation in diverse immune-mediated diseases. While IVIg actions are multifactorial, key therapeutic pathways rely on direct, structure-sensitive binding events between IgG and Fc receptors. In this study, we evaluated the chromatographic performance of receptor affinity columns utilizing FcγRIIIa and FcRn ligands to profile two IVIg products. Using pH-gradient elution on analytical nonporous resin (NPR) columns, we demonstrated that two receptors display distinct separation selectivities toward specific IgG Fc-glycoforms. Retention of the IgG glycoforms on the FcγRIIIa column was consistent with sensitivity to steric constraints imposed by both core fucosylation and galactosylation. Conversely, the FcRn column displayed a separation mechanism likely driven by charge heterogeneity. Despite the distal location of the glycosylation site relative to the binding interface with the IgG, the pH-gradient elution resolved sialylated subpopulations based on pI-mediated retention shifts. Both receptor columns yielded highly reproducible, product-specific fingerprints that enabled differentiation of the studied commercial preparations. Distinct fingerprints validate receptor affinity chromatography as a versatile tool for IVIg quality control. Furthermore, isolating distinct high-affinity subpopulations offers a preparative strategy for manufacturing next-generation IVIg therapeutics with tailored profiles. Specifically, FcγRIIIa-based enrichment could yield formulations with enhanced effector functions, such as increased antibody-dependent cellular cytotoxicity (ADCC), while FcRn-mediated fractionation could produce variants with a prolonged IgG half-life to extend biotherapeutic activity.
Keywords: affinity chromatography, immunoglobulin G, N-glycosylation, biotherapeutics, quality control
Presentation number: PP30
Abstract number: 119-ISABS-2026
AI-DRIVEN ECHOCARDIOGRAPHIC ASSESSMENT OF INTRACARDIAC MASSES
Đambić Klara6, Mešin Marko7, Brlek Petar1,2,3,4, Divanović Berina1, Bulić Luka1,2,3,5, Damjanović Ivan6, Hrvatin Nenad8,1, Primorac Dragan1,2,3,17,16,15,14,13,12,11,10,9
1St. Catherine Specialty Hospital, Zagreb, Croatia; 2International Center for Applied Biological Research, Zagreb, Croatia; 3School of Medicine, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia; 4Department of Molecular Biology, Faculty of Science, University of Zagreb, Zagreb, Croatia; 5Algebra Bernays University, Zagreb, Croatia; 6Department of Emergency Medicine, General Hospital Našice, Našice, Croatia; 7Institute of Emergency Medicine of Virovitica-Podravina county, Virovitica, Croatia; 8Department of Genomic Medicine, Clinical Hospital Center Rijeka, Rijeka, Croatia; 9Faculty of Dental Medicine and Health, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia; 10Eberly College of Science, The Pennsylvania State University, State College, Pennsylvania, United States of America; 11School of Medicine, University of Split, Split, Croatia; 12The Henry C. Lee College of Criminal Justice and Forensic Sciences, University of New Haven, New Haven, Connecticut, United States of America; 13Sana Kliniken Oberfranken, Coburg, Germany; 14School of Medicine, University of Rijeka, Rijeka, Croatia; 15School of Medicine, University of Mostar, Mostar, Bosnia and Herzegovina; 16National Forensic Sciences University, Gandhinagar, India; 17School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania, United States of America
klara.djambic97@gmail.com
Accurate detection and characterization of intracardiac masses remain a persistent challenge in cardiovascular imaging, primarily due to overlapping morphological and functional features among neoplastic and non-neoplastic entities. Recent advances in artificial intelligence (AI) are redefining the echocardiographic assessment of cardiac masses, with increasing impact on diagnostic accuracy and clinical decision making. Contemporary machine learning (ML), deep learning (DL), and multimodal AI approaches demonstrate that AI driven models, particularly convolutional neural networks and radiomics based techniques, can achieve high performance in detection, segmentation, and classification tasks. These methods also enable automated image acquisition and more standardized interpretation, reducing operator dependency. Furthermore, integration of echocardiographic data with complementary imaging modalities and clinical parameters enhances diagnostic precision and supports a more comprehensive evaluation framework. Despite these advances, several barriers continue to limit broader clinical adoption, including small and imbalanced datasets, heterogeneity of imaging data, limited external validation, and challenges related to model interpretability and generalizability. Taken together, current evidence positions AI assisted echocardiography as a key driver toward more objective, reproducible, and efficient cardiovascular diagnostics. Future progress will depend on large scale multicenter validation, development of explainable models, and seamless integration into routine clinical workflows.
Keywords: echocardiography, artificial intelligence, cardiac tumors, deep learning, machine learning
Presentation number: PP31
Abstract number: 90-ISABS-2026
LONG-TERM STABILITY ASSESSMENT OF IGG N-GLYCANS IN LYOPHILIZED PLASMA FOR GLYCOMICS
Fančović Matko1, Rapčan Borna1, Lauc Gordan1,2,1
1Genos Ltd., Zagreb, Croatia;2Faculty of Pharmacy and Biochemistry, University of Zagreb, Zagreb, Croatia
mfancovic@genos.hr
High-throughput glycomics relies on the availability of stable and reproducible samples for use as internal standards and in large-scale clinical cohorts. While previous literature has established the longitudinal stability of IgG N-glycans in regular frozen plasma, lyophilization offers superior advantages for global shipping and long-term storage logistics. This study evaluated a new lyophilization protocol to verify the structural integrity of the plasma glycome post-reconstitution. Human plasma samples were lyophilized and stored at room temperature (RT), +4°C, and -20°C, with stability assessed at baseline, 1 week, 1 month, 3 months, and 6 months. Analytical characterization was performed using Capillary Gel Electrophoresis with Laser-Induced Fluorescence (CGE-LIF). Our results demonstrate that IgG N-glycan profiles in lyophilized plasma remain stable for at least six months, with glycan concentrations and profiles matching those of non-lyophilized controls. While stability was maintained across all conditions, storage at -20°C provided the most robust results, yielding the highest peak intensities and superior analytical consistency. Conversely, samples stored at room temperature exhibited the highest Coefficients of Variation (CVs). These findings confirm that lyophilization is a highly effective preservation method that maintains the biochemical integrity of IgG N-glycans. Beyond its utility as a stable internal standard, this approach facilitates the simplified distribution of samples for multi-center studies, ensuring high-quality glycomic data while reducing the dependency on continuous cold-chain logistics.
Keywords: glycosylation, electrophoresis, plasma, lyophilization, immunoglobulin
Presentation number: PP32
Abstract number: 55-ISABS-2026
PHYSIOXIA-DRIVEN FUNCTIONAL AND SECRETOME PROFILING OF MENSTRUAL BLOOD-DERIVED STROMAL CELLS FOR PRECISION BIOMARKER DISCOVERY IN REPRODUCTIVE BIOCOMPATIBILITY AND INFERTILITY
Glusciukaite Kristina1, Pachaleva Jolita1, Piesiniene Lina2, Vidrinskaite Gabija1, Juskaite Pija1, Bernotiene Eiva1, Uzieliene Ilona1,2
1Innovative Medicine Centre, Vilnius, Lithuania; 2Nanodiagnostika ltd., Vilnius, Lithuania
kristina.glusciukaite@imcentras.lt
Infertility affects ~17.5% of couples worldwide, with up to 30% of cases remaining unexplained (uI), highlighting the need for precision medicine approaches and novel biomarker discovery. Human menstrual blood-derived mesenchymal stromal cells (MenSCs) provide a non-invasive model to study uterine biology, implantation, and partner-dependent reproductive compatibility. However, standard in vitro culture conditions fail to replicate the physiologically low oxygen environment of the endometrium. This study evaluates the effect of physioxia (5% O₂) versus normoxia (21% O₂) on MenSC functional, metabolic, and secretory profiles, and explores their relevance for identifying biomarkers linked to infertility and reproductive biocompatibility. MenSCs were assessed for proliferation (spectrophotometry), migration (holomonitor), immunophenotype and intracellular calcium (flow cytometry), metabolic activity (metabolism analyser), and secretome composition (multiplex assay). In parallel, extracellular vesicles (EVs) derived from menstrual blood, MenSCs, and semen are being investigated to characterize partner-specific molecular interactions (flow cytometry). Physioxia significantly enhanced MenSC proliferation, glycolytic capacity, and intracellular calcium signaling, while modulating surface marker expression (↓CD73; ↑CD105, CD90, CD13). Secretome analysis revealed increased production of growth factors and chemotactic mediators alongside reduced pro-inflammatory cytokines, indicating a microenvironment supportive of implantation. Ongoing multi-omic profiling (transcriptomics, proteomics) of EV cargo and MenSC–semen EV interactions aims to identify coordinated molecular signatures associated with altered reproductive compatibility. These findings support physioxia as a critical parameter for physiologically relevant in vitro modeling and highlight MenSC-derived systems as a platform for discovering couple-specific biomarkers.
Keywords: precision medicine, menstrual blood stromal cells, extracellular vesicles, unexplained infertility, multi-omics
Presentation number: PP33
Abstract number: 69-ISABS-2026
HEREDITARY TRANSTHYRETIN CARDIAC AMYLOIDOSIS IN A 50-YEAR-OLD FORMER PROFESSIONAL ATHLETE: A CASE REPORT
Jelinčić Petar1, Bulj Nikola1,2, Kovačić Đurđica1, Kovačić Jelena1,3, Primorac Dragan3,4,5,6,7,8,9,10,11,12,13
1School of Medicine, University of Zagreb, Zagreb, Croatia; 2University Hospital Centre “Sisters of Mercy”, Zagreb, Croatia; 3St. Catherine Specialty Hospital, Zagreb, Croatia; 4School of Medicine, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia; 5School of Medicine, University of Pittsburgh, Pittsburgh, PA, United States of America; 6School of Medicine, University of Split, Split, Croatia; 7International Center for Applied Biological Research, Zagreb, Croatia; 8Faculty of Dental Medicine and Health, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia; 9Eberly College of Science, The Pennsylvania State University, State College, PA, United States of America; 10Sana Kliniken Oberfranken, Coburg, Germany; 11School of Medicine, University of Rijeka, Rijeka, Croatia; 12Gandhinagar Campus, National Forensic Sciences University, Gandhinagar, India; 13The Henry C. Lee College of Criminal Justice and Forensic Sciences, University of New Haven, New Haven, CT, United States of America
pjelincic@student.mef.hr
Transthyretin amyloidosis is a progressive systemic disorder associated with noteworthy mortality and morbidity, commonly in older adults. TTR is a protein synthesized in the liver that functions in the transport of thyroxine and retinol, which consists of four monomers that assemble into a tetramer. It is characterized by structural and functional abnormalities caused by extracellular deposition of misfolded transthyretin protein as amyloid fibrils in multiple organs, with clinical manifestations typically dominated by cardiomyopathy and/or polyneuropathy. We present to you the case of a 50-year-old former professional athlete who was evaluated in the cardiology department with a family history of sudden cardiac death and amyloidosis. The patient complained of activity intolerance, irritable bowel syndrome, and symptoms of peripheral neuropathy. ECG findings included biatrial enlargement, increased myocardial wall thickness, and reduced longitudinal strain from base apex. Particularly, the apical strain was presented with a „cherry on top“ appearance, specific for cardiac amyloidosis. Serum protein electrophoresis showed no abnormalities, excluding the AL amyloidosis. Bone scintigraphy showed reduced bone uptake and increased cardiac uptake, consistent with ATTR amyloidosis. The diagnosis of hereditary ATTR amyloidosis was confirmed through genetic testing, which has also provided prognostic and therapeutic guidance. Following heart failure therapy showed modest symptomatic improvement and activity resumption. At one-year follow-up, disease progression and heart failure exacerbation led to the addition of tafamidis to the therapeutic regimen. Despite its rarity, cardiac amyloidosis should be considered in the differential diagnosis of a patient with heart failure, particularly when accompanied by a positive family history and systemic symptoms. Furthermore, this consideration allows for the timely introduction of treatment like tafamidis.
Keywords: transthyretin amyloidosis, cardiac amyloidosis, genetic testing, precision medicine, tafamidis
Presentation number: PP34
Abstract number: 110-ISABS-2026
ASSOCIATIONS BETWEEN INDOOR POLYCYCLIC AROMATIC HYDROCARBON EXPOSURE AND SERUM LIPIDOME PROFILES IN CHILDREN
Karlović Nina1, Gajski Goran2, Jakovljević Ivana2, Pehnec Gordana2, Turkalj Mirjana3,4,5, Banić Ivana3, Šarac Jelena1, Havaš Auguštin Dubravka1, Meščić Macan Andrijana1, Lovrić Mario1
1Institute for Anthropological Research, Zagreb, Croatia; 2Institute for Medical Research and Occupational Health, Zagreb, Croatia; 3Srebrnjak Children’s Hospital, Zagreb, Croatia; 4The School of Medicine, Catholic University of Croatia, Zagreb, Croatia; 5Faculty of Medicine, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia
nina.karlovic@inantro.hr
Exposure to indoor air pollutants, particularly polycyclic aromatic hydrocarbons (PAHs), represents a significant environmental health risk for children, yet the underlying biological mechanisms remain insufficiently understood. To quantify the association between indoor PAH exposure and the serum lipidome, an unsupervised principal component analysis (PCA) approach was used in 99 school-aged children from the Evidence Driven Indoor Air Quality Improvement (EDIAQI) project asthma cohort. Indoor air samples were analyzed for 11 PAHs, while serum lipid profiles were measured using high-throughput Nightingale NMR spectroscopy. Component scores were compared between high- and low-exposure groups (based on median splits) using Mann-Whitney U-tests and regression models adjusted for age, sex, and body mass index. PC1, PC2, and PC3 captured atherogenic lipoprotein burden, and HDL-versus-VLDL contrast, and triglyceride enrichment, respectively. Children with higher PAH exposure consistently showed higher PC1 and lower PC2 scores across 10 of the 11 PAHs, suggesting a subtle pro-atherogenic lipidome profile. The largest centroid separations occurred for benzo[g,h,i]perylene (ΔPC1 = 3.42) and fluoranthene (ΔPC1 = 3.21). A trend toward a positive association was found between benzo[a]pyrene and PC4 (β = 0.35, p = 0.067). Indoor PAH exposure appears to be modestly but consistently associated with a pro-atherogenic lipid profile in children.
Keywords: polycyclic aromatic hydrocarbons (PAHs), lipidomics, principal component analysis (PCA), indoor air pollution, environmental health
Presentation number: PP35
Abstract number: 42-ISABS-2026
FAMILIAL PORPHYRIA CUTANEA TARDA TYPE II ASSOCIATED WITH A NOVEL UROD I334T VARIANT AND HFE H63D
Kovačić Đurđica1, Vilibić Mirko1, Kovačić Jelena1,2, Jelinčić Petar1, Bulić Luka2,7,3, Brlek Petar2,3,6, Dulibić Mario1
1School of Medicine, University of Zagreb, Zagreb, Croatia; 2St. Catherine Specialty Hospital, Zagreb, Croatia; 3School of Medicine, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia; 4School of Medicine, University of Pittsburgh, Pittsburgh, PA, United States of America; 5School of Medicine, University of Split, Split, Croatia; 6Department of Molecular Biology, Faculty of Science, University of Zagreb, Zagreb, Croatia; 7Algebra Bernays University, Zagreb, Croatia; 8International Center for Applied Biological Research, Zagreb, Croatia; 9Faculty of Dental Medicine and Health, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia; 10Eberly College of Science, The Pennsylvania State University, State College, PA, United States of America; 11Sana Kliniken Oberfranken, Coburg, Germany; 12School of Medicine, University of Rijeka, Rijeka, Croatia; 13Gandhinagar Campus, National Forensic Sciences University, Gandhinagar, India; 14The Henry C. Lee College of Criminal Justice and Forensic Sciences, University of New Haven, New Haven, CT, United States of America
dkovacic@student.mef.hr
Porphyria is a group of rare metabolic disorders characterised by the buildup of porphyrins due to an abnormal heme biosynthesis pathway. Porphyria cutanea tarda (PCT) type II manifests as light-sensitive blistering cutaneous lesions and is caused by a heterozygous mutation in the UROD gene, which leads to a 50% reduction in uroporphyrinogen decarboxylase activity. However, symptoms develop when residual enzyme activity drops below a threshold of 25%. Some factors that contribute to such reduction of UROD activity are alcohol use, hepatitis C and genetic hemochromatosis. Genetic hemochromatosis is commonly caused by a homozygous C282Y mutation in the HFE gene. The H63D variant, which is prevalent in Mediterranean populations, does not cause significant iron overload on its own, but it may contribute to some abnormalities in its metabolism and, as such, is found to contribute to PCT onset and severity, especially when combined with other risk factors. We present the case of a 43-year-old male who was referred due to abnormal porphyrin and bilirubin findings. O/E the patient had vesicles and bullae, which had recently become less pronounced. His brother was diagnosed with porphyria, but genetic analysis was not conducted. The patient’s father also reported he’d had similar hand lesions in his thirties, which spontaneously resolved after a few years. Genetic testing was performed, and the report showed both a heterozygous variant of uncertain significance (VUS) c.1001T>C (I334T) in the UROD gene and a heterozygous pathogenic variant c.187C>G (H63D) in the HFE gene. To clarify the clinical significance of the VUS in the UROD gene, segregation analysis was recommended, and subsequently, the patient’s father and brother were tested. The same UROD VUS was detected in both. In this case, segregation of the UROD VUS in multiple family members provided evidence supporting its pathogenicity and highlighted its potential as a novel disease-causing variant in PCT type II.
Keywords: porphyria, porphyria cutanea tarda, HFE, UROD, whole genome sequencing
Presentation number: PP36
Abstract number: 92-ISABS-2026
PERSONALIZED HEALTH AND PERFORMANCE OPTIMIZATION IN PROFESSIONAL ATHLETES: INSIGHTS FROM ACE AND ACTN3 GENETIC POLYMORPHISMS
Lalovic Ana1, Dzuho Amra1, Smajlhodzic-Deljo Merima1, Softic Adna1, Asic Adna1
1Verlab Research Institute for Biomedical Engineering, Medical Devices and Artificial Intelligence, Sarajevo, Bosnia and Herzegovina
ana@verlabinstitute.com
Athletic performance arises from a complex interplay of genetic, physiological and environmental determinants. Among the most extensively investigated genetic contributors are polymorphisms in the ACE and ACTN3 genes, which have been linked to endurance and power-oriented phenotypes. This study aimed to examine the distribution of ACE and ACTN3 polymorphisms and their association with sport type in a cohort of 85 professional athletes from Bosnia and Herzegovina. Genomic DNA was analyzed using PCR-based genotyping methods, followed by appropriate statistical analyses. The ACE ID and ACTN3 RX genotypes were the most prevalent across the cohort, indicating a predominance of mixed performance profiles. The ACTN3 XX genotype was underrepresented, consistent with its lower frequency among elite athletes. A higher prevalence of the ACTN3 RR genotype and R allele was observed in power-oriented athletes, supporting its role in fast-twitch muscle performance. In contrast, ACE genotype distribution did not show significant variation between sport types. Notably, combined genotype analysis revealed that the ACTN3 RR and ACE DD “power” genotype combination was more frequent among power athletes, suggesting potential additive effects of multiple polymorphisms. However, logistic regression analysis did not identify any genetic markers as significant predictors of sport type, underscoring the multifactorial nature of athletic performance. Based on these findings, a preliminary framework for personalized training recommendations was developed, integrating genotype profiles with performance-related characteristics. These results support the potential of data-driven approaches, with future work focused on the development of artificial intelligence–based models to enhance predictive accuracy and enable scalable personalization in athlete training and performance optimization.
Keywords: ace, actn3, personalized health, polymorphisms, performance optimization
Presentation number: PP37
Abstract number: 100-ISABS-2026
INTEGRATED CLINICAL, BIOMECHANICAL AND MOLECULAR PROFILING OF GENERALIZED JOINT LAXITY AS A RISK FACTOR FOR ACL INJURY: PRELIMINARY RESULTS
Mešić Jana1,3, Jeleč Željko1,14, Brlek Petar1,2,3,12, Bulić Luka1,3,17, Škaro Vedrana2, Projić Petar2, Hrvatin Nenad1,13, Jozić Jakov1, Skelin Andrea1,15, Lauc Gordan15, Hanić Maja15, Jukić Igor16, Štimac Stjepan16, Primorac Dragan1,2,3,4,5,6,7,8,9,10,11
1St. Catherine Specialty Hospital, Zagreb, Croatia; 2International Center for Applied Biological Research, Zagreb, Croatia; 3School of Medicine, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia; 4Faculty of Dental Medicine and Health, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia; 5Eberly College of Science, The Pennsylvania State University, State College, PA, United States of America; 6School of Medicine, University of Split, Split, Croatia; 7The Henry C. Lee College of Criminal Justice and Forensic Sciences, University of New Haven, New Haven, CT, United States of America; 8Sana Kliniken Oberfranken, Coburg, Germany; 9School of Medicine, University of Rijeka, Rijeka, Croatia; 10School of Medicine, University of Pittsburgh, Pittsburgh, PA, United States of America; 11National Forensic Sciences University, Gandhinagar, India; 12Department of Molecular Biology, Faculty of Science, University of Zagreb, Zagreb, Croatia; 13Department of Genomic Medicine, University Hospital Centre Rijeka, Rijeka, Croatia; 14Department of Physiotherapy, University North, Varaždin, Croatia; 15Genos Glycoscience Research Laboratory, Zagreb, Croatia; 16Faculty of Kinesiology, University of Zagreb, Zagreb, Croatia; 17Algebra Bernays University, Zagreb, Croatia
jana.mesic@svkatarina.hr
This study aimed to systematically investigate clinical, biomechanical, and molecular correlates of generalized joint laxity as a potential risk factor for anterior cruciate ligament (ACL) injury. A total of 200 participants aged 18–50 years without prior ACL rupture were prospectively enrolled and stratified according to Beighton score (≥6 and <6). All participants underwent bilateral knee magnetic resonance imaging to exclude ACL pathology, comprehensive orthopedic examination, and objective biomechanical assessment, including instrumented Lachman testing and rotational stability analysis performed using the KiRA device. Peripheral blood and urine samples were obtained from all participants for targeted genetic, and glycomic analyses, as well as quantification of collagen degradation products, with analyses currently underway. Each group comprised 100 participants. The mean Beighton score was higher in females (5.86) compared to males (4.51). No statistically significant differences were observed between groups in anterior tibial translation. However, the Beighton <6 group demonstrated significantly lower rotational instability values, indicating greater rotational knee stability compared to the hypermobile cohort. These findings suggest that generalized joint laxity may preferentially influence rotational knee stability, thereby contributing to ACL injury susceptibility. Ongoing molecular analyses, including genetic profiling, glycan characterization, and assessment of collagen turnover biomarkers, are expected to further elucidate the underlying pathophysiological mechanisms and support the development of predictive, personalized risk stratification models.
Keywords: ACL, Beighton score, joint laxity, glycans, genetics
Presentation number: PP38
Abstract number: 91-ISABS-2026
GLYCOLYMPUS: A MODULAR FRAMEWORK FOR INTEGRATED PROCESSING AND VISUALIZATION OF GLYCOMICS DATA
Rapčan Borna1, Fančović Matko1, Lauc Gordan1,2,1
1Genos Ltd., Zagreb, Croatia; 2Faculty of pharmacy and biochemistry, University of Zagreb, Zagreb, Croatia
brapcan@genos.hr
Glycans are complex carbohydrate structures attached to proteins and lipids that play essential roles in cellular communication, immune response, and disease processes. Their comprehensive study, glycomics, is increasingly important for understanding biological systems and identifying biomarkers. However, glycomics data analysis typically involves multiple sequential steps, including raw data processing, electropherogram visualization, quantitative comparison, and report generation. These steps are often performed using separate tools, resulting in fragmented workflows, reduced reproducibility, and increased potential for user-induced errors. To address these challenges, we developed GlycOlympus, a modular desktop platform designed to integrate key components of glycan data analysis within a single, interactive environment. The platform consolidates essential functionalities, including input/output generation, electropherogram visualization, glycan data export, comparative analysis, and automated report generation. It supports workflows commonly used in capillary gel electrophoresis (CGE) and ultra-performance liquid chromatography (UPLC), enabling seamless transitions between different stages of analysis without the need for external tools. The application features a graphical interface that enables structured interaction with glycomics data across multiple stages of analysis, with an emphasis on consistent data organization and workflow continuity. Its modular design allows extensibility and facilitates the incorporation of additional analytical packages. By unifying multiple analytical steps into a coherent framework, GlycOlympus improves workflow efficiency and promotes reproducibility by reducing manual data handling and tool-switching. Overall, it provides a scalable, user-oriented solution for glycomics data analysis, contributing to more standardized and efficient analytical pipelines.
Keywords: glycomics, data interpretation, UHPLC, CGE, bioinformatics
Presentation number: PP39
Abstract number: 105-ISABS-2026
AI-DRIVEN IDENTIFICATION OF POLLUTION-ASSOCIATED TRANSCRIPTOMIC SIGNATURES IN ALZHEIMER’S DISEASE USING RNA-SEQ DATA
Softic Adna1,2, Salman Ali1, Luschi Alessio1, Becirovic Faruk3, Gurbeta Pokvic Lejla2, Iadanza Ernesto1
1Department of Medical Biotechnologies, University of Siena, Siena, Italy; 2Verlab Research Institute for Biomedical Engineering, Medical Devices and Artificial Intelligence, Sarajevo, Bosnia and Herzegovina; 3International University of Sarajevo, Sarajevo, Bosnia and Herzegovina
adna@verlabinstitute.com
Air pollution, particularly fine particulate matter (PM2.5 and PM10), has been increasingly associated with neurodegenerative diseases, including Alzheimer’s disease. However, the molecular mechanisms linking environmental exposure to neurodegeneration remain insufficiently understood. In particular, the integration of transcriptomic data with AI methods offers a promising approach for identifying disease-associated molecular signatures. In our study, we employed a machine learning-based framework to analyze publicly available RNA sequencing data from human hippocampal tissue (GEO accession: GSE161199). The dataset includes transcriptomic profiles from Alzheimer’s disease patients and age-matched controls. After preprocessing and normalization, feature selection was performed to identify the most informative genes associated with disease status, with a specific focus on genes previously reported to be responsive to oxidative stress, inflammation and environmental pollutants. Supervised ML models (Random Forest classifier and Support Vector Machine) were trained to distinguish between Alzheimer’s disease and control samples. Model performance was evaluated using cross-validation. To enhance interpretability, XAI methods, including SHAP, were applied to identify key genes contributing to model predictions. The analysis revealed a subset of genes involved in neuroinflammation, mitochondrial dysfunction and epigenetic regulation, which are also implicated in responses to air pollution exposure. These findings suggest a potential mechanistic link between environmental pollutants and transcriptomic alterations observed in neurodegenerative diseases. This study demonstrates the feasibility of integrating AI with publicly available multi-omics datasets to uncover environmentally relevant molecular signatures in Alzheimer’s disease. The proposed approach highlights the potential of AI-driven models in identifying early biomarkers associated with environmental risk factors.
Keywords: air pollution, PM, Alzheimer’s disease, ecogenotoxicology, neurodegenerative disorders
Presentation number: PP40
Abstract number: 37-ISABS-2026
OMICS APPROACH FOR DISCOVERING IMMUNE CANDIDATE GENES IN MULTIPLE SCLEROSIS
Stanković Matić Ivana1, Peterlin Borut2, Turk Aleksander2, Starčević Čizmarević Nada1
1Department of Medical Biology and Genetics, Faculty of Medicine, University of Rijeka, Rijeka, Croatia; 2Clinical Institute of Genomic Medicine, University Medical Centre Ljubljana, Ljubljana, Slovenia
ivanasm@uniri.hr
Multiple sclerosis (MS) is a complex immune-mediated disorder with a polygenic architecture. Identification of causal genes remains challenging due to the indirect nature of genome-wide association studies (GWAS). We propose an integrative multi-omics framework for prioritizing immune-relevant candidate genes and supporting targeted panel design. Candidate genes are derived from published literature, GWAS datasets, and validated susceptibility variants from the International Multiple Sclerosis Genetics Consortium (IMSGC). Gene prioritization was performed using locus-to-gene (L2G) scores from Open Targets Genetics, integrating genetic distance, eQTL/sQTL colocalization, chromatin interaction data, and predicted functional impact of coding variants. Immune relevance is refined using curated pathway databases (Reactome, KEGG, ImmPort). Multi-layered evidence from transcriptomic, epigenomic, and proteomic datasets is integrated into a unified scoring framework to stratify genes into high- and medium-confidence categories numbering 100-150 genes in Next Generation Sequencing (NGS) panel. The study cohort includes 300 MS patients, stratified into 100 familial cases, 150 sporadic cases, and 50 individuals from the Gorski Kotar region, which represents a region of particular epidemiological interest for MS research in Croatia. The integrative framework is expected to prioritize immune-related genes supported by convergent genetic and functional evidence. High-confidence candidates are anticipated to converge on key biological pathways, including cytokine signaling, lymphocyte activation, and neuroinflammatory regulation. This approach enables systematic gene prioritization and supports the design of a targeted NGS panel, facilitating downstream sequencing and comparative analysis across familial, sporadic, and regionally enriched MS subgroups.
Keywords: genome-wide association studies, immunology, multi-omics, multiple sclerosis, next generation sequencing panel
Presentation number: PP41
Abstract number: 26-ISABS-2026
LIFESTYLE AND DIETARY INFLUENCES ON GUT MICROBIOTA DIVERSITY AND COMPOSITION IN A HEALTHY CROATIAN COHORT
Šarac Jelena1,2, Havaš Auguštin Dubravka1, Šunić Iva1, Bočkor Luka1, Novokmet Natalija1, Mrdjen- Hodžić Rafaela1, Michl Kristina3,4, Wicaksono Wisnu Adi3, Černava Tomislav3,5, Lovrić Mario1,2,6, Marjanović Damir1,2,7
1Centre for Applied Bioanthropology, Institute for Anthropological Research, Zagreb, Gajeva 32, Croatia; 2Faculty of Biotechnology and Drug Development, University of Rijeka, Rijeka, Croatia; 3Institute of Environmental Biotechnology, Graz University of Technology, Graz, Austria; 4Department Life Science Engineering, University of Applied Sciences Technikum Wien, Vienna, Austria; 5School of Biological Sciences, Faculty of Environmental and Life Sciences, University of Southampton, Southampton, United Kingdom of Great Britain and Northern Ireland; 6Faculty of Food Technology Osijek, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia; 7Burch University, Sarajevo, Bosnia and Herzegovina
jsarac@inantro.hr
The human gut microbiota plays a key role in health and is influenced by different environmental factors. This study examined associations between microbial diversity and composition in the gut with lifestyle and dietary factors in a healthy Croatian cohort (N=60). 16S rRNA gene sequencing was performed on the Illumina platform, following standard procedures. Genetic data were processed in R using the phyloseq package. Alpha diversity was analyzed using linear models and Kruskal–Wallis tests, while beta diversity (Bray–Curtis) was assessed using PERMANOVA. Genus-level relative abundance profiles were generated to characterize microbial composition. Beta diversity analysis revealed significant differences in microbial community structure associated with gender (p=0.024), physical activity (p=0.013), bloating frequency (p=0.002), and weekly wine consumption (p=0.047). In contrast, only age showed a marginal association with alpha diversity. Wine consumption was most significantly associated with variation in Faecalibacterium (p=0.011) and Ruminococcus (p=0.007), physical activity with [Eubacterium] hallii group (p=0.004, q=0.031) and Bifidobacterium (p=0.003, q=0.031), and meat consumption with Anaerostipes (p=0.003, q=0.025) and [Eubacterium] hallii group (p<0.001, q=0.005). Soft drink consumption showed limited associations. Relative abundance profiles showed consistent but modest shifts across lifestyle categories, including trends toward higher proportions of short-chain fatty acid–producing genera (e.g., Faecalibacterium, Roseburia) with increased physical activity and lower processed beverage intake. These taxa are commonly considered beneficial due to their role in short-chain fatty acid production, maintenance of gut health and healthy aging in general. These findings suggest that while lifestyle and dietary factors may not substantially alter microbial richness, they are associated with compositional shifts in specific bacterial taxa.
Keywords: gut microbiota, lifestyle, diet, health, bacterial taxa
Presentation number: PP42
Abstract number: 4-ISABS-2026
INTEGRATIVE EPIGENETIC AND TRANSCRIPTOMIC PROFILING REVEALS DNA METHYLATION–GENE EXPRESSION RELATIONSHIPS IN IDIOPATHIC ALS
Teixeira da Silva Hucke Andre1, Zhao Tianying2, Hoffman Mariah2, Landman Bennett1, Belzil Veronique2
1Vanderbilt University, Nashville, Tennessee, United States of America; 2Vanderbilt University Medical Center, Nashville, Tennessee, United States of America
andre.hucke@vanderbilt.edu
Epigenetic modifications capture gene–environment interactions and reflect biological processes associated with various diseases and aging. These features offer a potential molecular record of tissue injury accessible through peripheral biofluids. In idiopathic amyotrophic lateral sclerosis (ALS) cases, noninvasive biomarkers that reflect neuronal pathology remain limited. Recent studies have explored the circulating cell-free DNA (cfDNA) methylome, linking epigenetic alterations to neurodegenerative pathways, such as endocytosis. However, the cfDNA methylome in idiopathic ALS, derived from dying neurons and glia, and its relationship to brain transcriptional dysregulation remain unclear. To address this gap, we generated high-resolution DNA methylation profiles from cfDNA and matched genomic DNA (gDNA) in 48 blood samples, including 24 individuals with idiopathic ALS and 24 controls. Enzymatic methylation sequencing was employed to preserve DNA integrity and maximize signal recovery from low-input cfDNA. Differentially methylated regions were identified using a tiling window approach (q ≤ 0.01; ≥10% methylation difference). Comparisons between cfDNA- and gDNA-derived methylation profiles were used to distinguish injury-associated epigenetic signals from stable blood cell–intrinsic methylation states, and CelFiE-based deconvolution was applied to infer cellular contributors to cfDNA. Circulating methylation features were integrated with brain-derived bulk and single-nucleus RNA sequencing data to prioritize pathways relevant to neurodegeneration and metabolic dysregulation. This analysis identified concordant methylation and transcriptional alterations near key metabolic regulators, including ACSS2, PCCB, and ABAT. Epigenetic age estimates derived from cfDNA were not directly comparable to clocks calculated from matched gDNA, highlighting limitations of existing epigenetic aging models when applied to cfDNA and underscoring the need for cfDNA-optimized clocks.
Keywords: dna methylation, amyotrophic lateral sclerosis, epigenetic aging, multi-omics integration
Presentation number: PP43
Abstract number: 9-ISABS-2026
PHYSICAL ACTIVITY AND IGG N-GLYCOMIC PROFILE: A CROSS-SECTIONAL STUDY AMONG MEDICAL STUDENTS
Vidović Stipe1,2, Maduna Ivanka1,7, Mešin Marko1, Fančović Marko3, Hanić Maja3, Heffer Marija1,4, Lauc Gordan3, Zibar Lada1,5,6
1Faculty of Medicine Osijek, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia; 2Clinic for Eye Diseases, University Hospital Centre Osijek, Osijek, Croatia; 3Genos Glycoscience Research Laboratory, Zagreb, Croatia; 4Department of Medical Biology and Genetics, Faculty of Medicine Osijek, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia; 5Department of Pathophysiology, Faculty of Medicine Osijek, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia; 6Department of Nephrology, University Hospital Merkur, Zagreb, Croatia; 7Health Center of Osijek-Baranja County, Osijek, Croatia
stipevidovic1@gmail.com
The aim of this study was to examine physical activity (PA) among medical students, to measure their IgG N-glycan composition and determine differences in the IgG N-glycan profile according to PA category, and to explore associations between specific types of PA and IgG N-glycan traits. This cross-sectional study was conducted in December 2023 among first- and second-year medical students at the University of Osijek, Croatia. PA was assessed using the International Physical Activity Questionnaire – Short Form (IPAQ-SF) questionnaire. IgG was isolated from plasma, and its N-glycan composition was analyzed by capillary gel electrophoresis (27 IgG N-glycan peaks, and eight derived glycan traits representing shared structural features). Results: A total of 79 students (23 males), median age 20 years (interquartile range 19 – 20), completed the questionnaire. According to the IPAQ-SF, 23 participants were classified as health-enhancing physical activity (HEPA)-active, while 56 were categorized as non–HEPA-active. After Benjamini–Hochberg false discovery rate (FDR) correction, no significant differences were found in IgG N- glycan peaks or traits between groups. Initial correlations were observed between vigorous PA and monogalactosylated glycan traits (G1) (ρ = −0.24, P = 0.034), total PA and digalactosylated glycan traits (G2) (ρ = −0.31, P = 0.005) and G1 (ρ = 0.25, P = 0.024), and walking and G2 (ρ = −0.23, P = 0.046), but none remained significant after FDR correction (P ≥ 0.218). PA in medical students was not prevalent. No significant differences or associations between PA and the IgG N-glycan profile were identified after multiple testing correction in this young, homogeneous population. Larger longitudinal and interventional studies with objective PA assessment are needed. In this context, artificial intelligence–based multi-omics integration may help identify complex PA–IgG glycosylation interactions.
Keywords: glycomics, immunoglobulin G, medical students, N-glycans, physical activity
Published: June 16th, 2026.
Copyright: © 2026 Authors of Ai-Powered Multi-Omics Integration For Precision Medicine section. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
