Dengue virus (DENV) serotyping is essential for epidemiological surveillance, clinical risk assessment, and vaccine evaluation, as the four dengue serotypes differ in pathogenicity, immune interactions, and population dynamics. Existing subtyping methods largely rely on sequence alignment and phylogenetic inference, which can be computationally intensive and unreliable for short, fragmented, or error-prone sequences commonly generated in diagnostic and surveillance settings. There is a need for fast, alignment-free serotyping approaches that maintain high accuracy across heterogeneous sequence lengths while remaining scalable, transparent, and suitable for real-world diagnostic inputs. We demonstrate that compact 3-mer composition features are sufficient for highly accurate dengue virus serotyping when coupled with a lineage-aware Random Forest classification framework. Using 64 normalized 3-mer frequency features per sequence with ambiguity masking and enforcing strict cluster-aware validation at both 99% and 95% nucleotide identity thresholds, our approach achieved near-perfect accuracy and macro-F1 scores on held-out internal test sets. To further ensure independence, external validation datasets were filtered to remove exact sequence matches and any sequences sharing ≥99% or ≥95% nucleotide identity with internal data. On these strictly independent external datasets, the model maintained 100% accuracy and macro-F1 performance, confirming robust generalization beyond database redundancy. Robustness analyses showed stable performance under contiguous sequence truncation down to 300 bp and in the presence of ambiguous nucleotides, indicating resilience to realistic diagnostic inputs. These results demonstrate that a lightweight, alignment-free, machine learning approach can rival alignment-dependent methods while maintaining strict lineage-aware evaluation controls. The proposed framework combines high predictive accuracy, probabilistic reliability, computational efficiency, and reproducible validation design, making it well suited for large-scale genomic surveillance, rapid pre-screening, and diagnostic decision-support applications.

Identifying the areas where WNV circulates before the occurrence of human infection cases is of a crucial importance for disease risk reduction and epidemic prevention. Screening field-collected mosquitoes can reveal WNV circulation two to five weeks before the first human cases occur. Timely surveillance for the early detection of WNV circulation is the primary measure and a prerequisite for the successful implementation of prevention and control strategies. To define the exact timing and the conditions that play an essential role in appearance and multiplication of WNV in vector mosquitoes and the onset of natural WNV circulation during a season, a comprehensive study on WNV presence in vector mosquitoes was conducted at 13 localities in the Vojvodina Province of Serbia during 2024. From April 20th to September 26th, 2024, eggs, larvae, and female adults of Culex pipiens, the most competent vector of WNV in the region, were collected on a weekly basis. In addition, to each CO₂ baited mosquito trap for adults' collection, a commercially available FTA card soaked with liquid honey was fixed to the inner side of the trap nets in order to collect the saliva of captured adults. During the sampling environmental data collection (relative humidity and temperature) at each collection site was conducted using data loggers. Environmental and biological data sets were used for modelling and risk prediction in an early warning system. In the laboratory, Cx. pipiens egg rafts, larvae, and adult mosquitos were individually processed (mechanically disrupted) and prepared for molecular detection of WNV. In total, 7340 individual samples were tested (207 FTA cards, 5875 adults female Cx. pipiens mosquitoes, 19 eggs rafts, 1199 larvae and 40 pupae). In order to analyse the circulating biotypes of Cx. pipiens, and to determine whether the molestus, pipiens biotypes and their hybrids are present, as well as to determine the biotypes of WNV-positive mosquitoes the chosen samples (in total 605 samples) were sent to another project partner institution (IRTA CReSA, Barcelona, Spain). Preliminary results confirmed that the first detection of WNV was in an adult mosquito Cx. pipiens in mid-May 2024, and only one week later WNV was detected on the FTA card. WNV was detected only in adult mosquitoes of Cx. pipiens and on FTA cards; none of the egg, pupa or larval samples tested were positive for WNV. The highest number of WNV-positive samples was recorded in August and September 2024. Out of in total 5875 tested adult Cx. pipiens mosquitoes, WNV was detected in 10 individuals from 5 out of 13 sampled locations. Also, WNV RNA was detected in 4 out of 207 tested FTA cards from 4 out of 13 sampled locations. In 2 locations WNV presence was detected in both adult mosquito and FTA cards, and in only one occasion WNV was detected at the same time in the adult mosquito and the FTA card in the same trap. In total, WNV presence was detected in 7 out of 13 sampled locations. Valuable and practical conclusions will be drawn when all additional sample testing is completed and after incorporating other systematically collected environmental data during the sampling period (meteorological data, habitat type of sampling site, etc.). The ultimate goal is to develop a prediction model that is expected to provide the necessary data to design an effective monitoring system for early detection of WNV circulation in an area.

Extensive pig production represents a significant segment of livestock production in Serbia. Due to the continuous presence of African swine fever (ASF) in Serbia since 2019, with the predominant number of outbreaks diagnosed on this type of holdings, the level of biosecurity measures applied on them is of particular importance. However, the impact of sociodemographic and structural characteristics of holdings on the level of applied biosecurity measures has not been sufficiently studied. The aim of this study was to assess the level of biosecurity measures on extensive pig holdings in Vojvodina using the Biocheck.UGent tool, and to examine the relationship between biosecurity level and farmer experience and holding size. A cross-sectional study was conducted in 2026 and included 26 rural holdings and small extensive-type farms in 12 municipalities within 4 districts in Vojvodina. A total of 3,592 pigs were recorded, with an average of 138 animals per holding (range 12-600). The average working experience of farmers in pig production was 25.5 years (range 5-50). Biosecurity was evaluated using the Biocheck.UGent questionnaire. Pearson correlation analysis was applied to examine the relationship between biosecurity scores, number of animals, and years of farmer experience. The mean overall biosecurity score was 65.5% (range 35-83%), with external biosecurity (64.0%) being lower than internal biosecurity (68.8%). Correlation analysis showed no statistically significant relationship between years of farmer experience and overall biosecurity score (r=0.078; p>0,05), nor between experience and external (r=0.158; p>0,05) or internal biosecurity (r=-0.078; p>0,05). Similarly, no significant correlation was found between the number of animals per holding and overall biosecurity score (r=-0.137; p>0,05). The results indicate that neither long-term farmer experience nor holding size significantly determine the level of biosecurity. This finding suggests that empirical experience without structured education is insufficient for adequate implementation of biosecurity measures. Implementation of educational and advisory programs represents a necessary prerequisite for improving biosecurity on extensive-type holdings in Vojvodina.