Background: The significance of lactylation in human disease is well-established, but its specific involvement in age-related macular degeneration has yet to be fully elucidated. Objective: This study aims to explore lactylation-related genes in age-related macular degeneration. Methods: Gene expression data from retinal pigment epithelium-choroid samples obtained from age-related macular degeneration patients were collected and integrated from the GSE29801 and GSE50195 datasets. Differential analysis was conducted, followed by the application of machine learning techniques to identify key lactylation-related hub genes implicated in age-related macular degeneration. Subsequently, these 7 identified lactylation-related hub genes underwent correlation analysis, immune cell infiltration analysis, receiver operating characteristic analysis, and unsupervised clustering to further investigate their potential roles in age-related macular degeneration. In addition, an assessment was conducted on the potential role of seven hub genes in patients with age-related macular degeneration. Results: Through the analysis of differentially expressed genes, a total of 68 lactylation-related genes were identified as overlapping. Among these, TOP2B, HMGN2, SRRM2, HNRNPH1, SF3A1, HIST1H1C, and HIST1H2BD were recognized as hub genes. Conclusion: The present study has thus identified seven lactylation-related genes that may have significant implications in the pathogenesis of age-related macular degeneration. These findings offer novel perspectives for further exploration of age-related macular degeneration and its underlying mechanisms.

Haemophilia ’A’, an X-linked genetic disorder stemming from the deficiency of coagulation factor VIII, exhibits varying severity levels: mild with infrequent bleeding, moderate involving bleeding after minor injuries, and severe leading to spontaneous bleeding, posing risks of haemophilic arthropathy and physical disability. Tailoring treatments for these levels is crucial, with past studies suggesting that factors beyond mutations, such as mutation location, contribute to severity. The SNP analysis is vital for accurately assessing disease severity. The imbalanced nature of Haemophilia SNP data in Mild, Moderate, and Severe Cases may compromise the accuracy of classifier predictions.To address this imbalance, propose a LOF, ADASYN, and NCL (LAN) hybrid to balance PSM-processed SNP data. Employing the P-LAN methodology for SNP data analysis, a thorough evaluation encompasses classical supervised machine learning algorithms and proposed innovative embedded classifiers CAT-SVM. Validating the proposed CAT-SVM model through a robust 10-fold cross- validation scheme reveals enhanced mean accuracy across classifiers, surpassing previous benchmarks with up to 91.36% accuracy. Notably, when applied to imbalanced test data related to severity levels, the model demonstrates a commendable test accuracy of up to 85%, indicating a substantial improvement in predictive capability compared to earlier studies.Comprehensive in-silico analysis reveals the transitional behaviour observed within the moderate class, suggesting the need for a unified classification for ’Mild’ and ’Moderate’ severity. In the future, this study can be utilized to enhance severity prediction models in the context of mutation-based disorder and extend to other imbalanced datasets. It will be ultimately pave way for personalized treatment of individuals based on genotype.