Journal of Applied Genetics最新文献

{"title":"Integrating speed breeding and genomic tools to accelerate genetic gain in chilli (Capsicum annuum L.).","authors":"Sahana Patage, Usha Nandhini Devi H, S Praneetha, M Djanaguiraman, N Premalatha","doi":"10.1007/s13353-026-01100-8","DOIUrl":"https://doi.org/10.1007/s13353-026-01100-8","url":null,"abstract":"<p><p>Traditional breeding of chilli pepper (Capsicum annuum L.) is constrained by long generation intervals, limiting the rate of annual genetic gain. Recent advances in predictive breeding, genomic prediction, and image-based high-throughput phenotyping have created new opportunities for integrated breeding pipelines in Capsicum annuum. Speed breeding (SB) shortens the generation cycles through extended photoperiods and optimized light regimes, enabling rapid recycling of breeding populations. This review synthesizes the integration of SB with doubled haploid (DH), genomic selection (GS), and high-throughput phenotyping (HTP) to propose a crop-specific predictive breeding framework for chilli pepper improvement. Under optimized SB protocols, flowering time and seed maturation can be substantially accelerated, while DH technology enables rapid fixation of elite recombinants despite persistent genotype-dependent limitations. Genomic selection improves prediction accuracy for complex quantitative traits, including stress tolerance and disease resistance, whereas HTP platforms facilitate rapid and non-destructive phenotypic evaluation. Recent Capsicum studies further demonstrate the importance of multi-environment genomic prediction for addressing genotype-by-environment interactions in breeding programs. However, challenges including low DH efficiency, environmental sensitivity, and limited transferability between controlled and field conditions remain significant constraints. Overall, the integrated SB-DH-GS-HTP framework provides a scientifically grounded strategy for improving breeding efficiency and accelerating the development of climate-resilient chilli cultivars.</p>","PeriodicalId":14891,"journal":{"name":"Journal of Applied Genetics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2026-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148677990","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The impact of FTO and ADRB3 gene variants, physical activity and proper diet on weight gain and perinatal outcomes in pregnant women with obesity.","authors":"Urszula Mantaj, Łukasz Adamczak, Barbara Więckowska, Rafał Sibiak, Ewa Wender-Ożegowska","doi":"10.1007/s13353-026-01104-4","DOIUrl":"https://doi.org/10.1007/s13353-026-01104-4","url":null,"abstract":"<p><p>This study aimed to evaluate the effects of FTO (rs9939609) and ADRB3 (rs4994) gene variants, dietary intervention, and physical activity on gestational weight gain and perinatal outcomes in women with obesity and early gestational diabetes (GDM). In this prospective study, 115 pregnant women with obesity and early GDM were enrolled. All participants received standardised dietary counselling and recommendations regarding physical activity. Gestational weight gain, biochemical parameters, physical activity levels, genetic variants, and maternal and neonatal outcomes were assessed throughout pregnancy. Appropriate gestational weight gain was achieved by 73% of participants. Both earlier diagnosis of GDM (11 vs 14 weeks, p = 0.04) and initiation of dietary intervention (12 vs 15 weeks, p = 0.01) were observed in the group with recommended gestational weight gain. Women with EGWG had significantly higher glycated haemoglobin (HbA1c) levels in the second (5.03% vs 5.22%, p = 0.02) and third (5.33% vs 5.55%, p = 0.02) trimesters. Unfavourable FTO genotypes were associated with higher first-trimester HbA1c concentrations (p = 0.04) and elevated triglyceride levels throughout pregnancy (p < 0.05) but were not independently associated with gestational weight gain. Higher physical activity in each trimester, was significantly associated with appropriate gestational weight gain regardless of genetic predisposition (p < 0.001). Appropriate gestational weight gain was associated with a lower risk of neonatal macrosomia (2% vs 4%, p = 0.04). Among women with obesity and early GDM early dietary intervention and sustained physical activity were more strongly associated with appropriate gestational weight gain than the studied genetic variants. Appropriate gestational weight gain was associated with a lower risk of neonatal macrosomia.Trial registration number and date of registration: NCT04924738, 14/06/2021.</p>","PeriodicalId":14891,"journal":{"name":"Journal of Applied Genetics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2026-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148673855","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integration of graphical and diallel analyses for identifying elite bread wheat genotypes for grain yield and its yield-related traits.","authors":"Khaled F M Salem, Essam F El-Hashash, Mohamed Ali Abdelsatar, Karima M El-Absy, Ulkar Ibrahimova, Hassan A Mesbah, Amged El-Harairy, Amira A Ibrahim","doi":"10.1007/s13353-026-01101-7","DOIUrl":"https://doi.org/10.1007/s13353-026-01101-7","url":null,"abstract":"<p><p>Improving grain yield and its components through a comprehensive understanding of genetic parameters continues to be a major challenge in wheat breeding, owing to the inherent complexity of simultaneously enhancing multiple agronomic traits. In the present study, the inheritance pattern of grain yield and its related traits in seven bread wheat (Triticum aestivum L.) genotypes, along with their F₁ progeny, was investigated using Hayman's numerical and graphical diallel analyses within the framework of a randomized complete block design with three replications. Morley-Jones analysis revealed significant differences (p < 0.05 or 0.01) for grain yield and most traits. Inheritance was predominantly governed by both additive and dominant gene effects, such that non-additive action, particularly overdominance, played a fundamental role in grain yield per plant (GYP) and most traits. The parental genotypes exhibited a higher frequency of dominant alleles compared to recessive alleles, and favorable alleles outnumbered unfavorable ones. The results of regression analysis showed the absence of non-allelic interactions for the traits plant height (PH), spike length (SL), spike weight (SW), number of grains per spike (NGS), spike grain weight (SGW), and number of spikelets per spike (NSP); while the presence of epistatic interactions was detected for number of tillers (NT), number of spikes (NS), and GYP. Based on low Wr + Vr, it is inferred that dominant alleles primarily control SD, SGW, SW, and SL. Also, the results of graphical analysis showed that the partial dominance pattern prevailed for PH, SL, NT, NS, and GYP. In contrast, overdominance was observed for SGW, SD, NGS, and SW, whereas the trait number of NSP exhibited a complete dominance pattern. Genotypes P6 and P7 carried more recessive alleles, whereas P2 and P3 had higher dominant allele frequencies, suggesting potential hybridization to develop superior lines. GGE biplot analysis identified P2 (Sakha 93) as the top performer for SD, NS, NT, and NSP, while P2, P3, and P5 combined high stability and grain yield, positioning them near the ideal genotype.</p>","PeriodicalId":14891,"journal":{"name":"Journal of Applied Genetics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2026-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148630303","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A novel multi-epitope mRNA vaccine against Escherichia coli using an in silico reverse vaccinology approach.","authors":"Shakeela Parveen, Javeria Abbasi, Muhammad Hussain, Muhammad Farhan Khan, Muhammad Tayyab, Amina Parveen, Laiba Shafique","doi":"10.1007/s13353-026-01102-6","DOIUrl":"https://doi.org/10.1007/s13353-026-01102-6","url":null,"abstract":"<p><p>The adaptable nature of Escherichia coli, shifting from benign to virulent forms, poses a major healthcare challenge due to rising multidrug resistance and hypervirulence. In this study, using an in silico reverse vaccinology approach, we designed a multiepitope mRNA vaccine candidate targeting the Shiga toxin 1 (Stx1) of Shiga toxin-producing E. coli (STEC). Two cytotoxic T-Lymphocyte (CTL), three Helper T-Lymphocyte (HTL), and two Linear B-Lymphocyte (LBL) epitopes were identified and selected through stringent computational filtering based on high antigenicity, non-allergenicity, and non-toxicity. These epitopes were joined using optimized linkers (EAAK, AAY, GPGPG, KK) and the adjuvant PefE to form a stable 184-residue multi-epitope construct. Physicochemical profiling predicted a molecular weight of 19969.45 Da, an antigenicity score of 0.9278 (VaxiJen v.2.0), a basic isoelectric point (pI) of 10.31, and structural stability. Structural analysis revealed 21.74% alpha-helical content. Molecular docking exhibited strong binding with human Toll-like receptor 2 TLR2 and TLR4, with weighted energy scores of - 1096.2 and - 1110.5 kcal/mol, respectively, mediated by extensive hydrogen-bonding and salt-bridge networks. In silico immune simulation projected a strong clonal expansion of B-cells, active T-helper and cytotoxic T-cells, and higher cytokine production (IFN-γ, IL-10). Codon Adaptation Index (CAI) of 0.87 and a stable mRNA secondary structure were predicted with a minimum free energy (MFE) of - 204.10 kcal/mol. While these computational findings provide immunogenic potential, they represent a predictive hypothesis. Subsequent in vitro synthesis and in vivo testing in animal models are essential to validate its safety, immunogenicity, and protective efficacy.</p>","PeriodicalId":14891,"journal":{"name":"Journal of Applied Genetics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2026-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148620092","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integrated bioinformatics and experimental analysis identify apoptosis- and pyroptosis-related hub genes as candidate diagnostic biomarkers in Alzheimer's disease.","authors":"Yawen Cai, Yifan Chen, Guiqin Huang, Menghui Ren, Yuhui Chai, Cong Gai, Xi Huang, Tianhua Yan","doi":"10.1007/s13353-026-01095-2","DOIUrl":"https://doi.org/10.1007/s13353-026-01095-2","url":null,"abstract":"<p><p>Apoptosis and pyroptosis-mediated neuronal death represent major pathogenic mechanisms underlying Alzheimer's disease (AD). Given the potential crosstalk between these two forms of cell death, investigation of a single death pathway may be insufficient to identify robust diagnostic biomarkers for AD. Therefore, this study aimed to explore hub genes involved in both apoptosis and pyroptosis as potential diagnostic biomarkers for AD. First, 23 common cell death-related genes (CDRGs) were identified through bioinformatic analysis. Functional enrichment analyses using GO, KEGG, and GeneMANIA revealed significant associations between AD and biological processes including apoptosis, pyroptosis, and neuronal death. Subsequently, machine learning algorithms combined with ROC curve analysis identified CASP3, IL1B, NLRP3, and PYCARD as candidate biomarkers with potential diagnostic and therapeutic implications for AD. These findings were further validated by the in vitro experiments, which confirmed that the expression levels of these four biomarkers were consistent with the predicted results. Additionally, in patients with AD, CASP3, IL1B, NLRP3, and PYCARD were negatively correlated with macrophages. Collectively, these results suggest that the identified biomarkers may co-regulate apoptosis and pyroptosis through macrophages, thereby contributing to the pathogenesis of AD.</p>","PeriodicalId":14891,"journal":{"name":"Journal of Applied Genetics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2026-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148535860","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Impaired strigolactone biosynthesis and signalling are associated with promoter-associated CHH hypermethylation in Arabidopsis thaliana.","authors":"Agnieszka Brąszewska, Marek Marzec","doi":"10.1007/s13353-026-01099-y","DOIUrl":"https://doi.org/10.1007/s13353-026-01099-y","url":null,"abstract":"<p><p>Strigolactones are plant hormones that regulate diverse developmental processes and stress responses, but their potential role in epigenetic regulation remains poorly understood. Previous transcriptomic analyses of strigolactone mutants in various species revealed extensive transcriptional reprogramming, suggesting destabilisation of higher-order regulatory mechanisms rather than changes in discrete gene networks. Here, we investigated whether disruption in strigolactone biosynthesis or perception is associated with altered DNA methylation patterns. Whole-genome bisulfite sequencing was performed on 7-day-old seedlings of Arabidopsis thaliana wild-type and the strigolactone biosynthesis mutant max3 and the perception mutant d14. A comparative analysis identified differentially methylated regions in CG, CHG, and CHH sequence contexts. While CG methylation changes showed no directional bias, non-CG contexts exhibited a pronounced enrichment of hypermethylated regions in both mutants, with the strongest effect observed in the CHH context. CHH hypermethylation was preferentially enriched at promoter-associated regions, primarily overlapping non-coding and pseudogene loci and showed substantial overlap between max3 and d14, indicating a shared epigenetic response in both mutants. Gene-body-associated regions displayed a similar but weaker pattern. These results demonstrate that impairment of the strigolactone pathway is consistently associated with selective CHH hypermethylation rather than global methylation changes. Because CHH methylation in plants is commonly associated with RNA-directed DNA methylation, the observed pattern is consistent with altered RdDM-associated regulation; however, direct involvement of the RdDM machinery remains to be established. This study provides a genome-wide framework for exploring how strigolactones may be associated with epigenetic regulation in plants.</p>","PeriodicalId":14891,"journal":{"name":"Journal of Applied Genetics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2026-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148535857","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Identification of novel variants in SSX1, CPNE1, and SPTBN5 in men with oligoasthenoteratozoospermia using whole-genome sequencing.","authors":"Sara Sadeghzadeh, Agnieszka Malcher, Razieh Ebrahimi Askari, Rim Ibrahim, Fatemeh Sefid, Krzysztof Pastuszak, Aashish Srivastava, Mohammadreza Dehghani, Serajoddin Vahidi, Marzieh Lotfi, Nasrin Ghasemi, Maciej Kurpisz","doi":"10.1007/s13353-026-01089-0","DOIUrl":"https://doi.org/10.1007/s13353-026-01089-0","url":null,"abstract":"<p><p>Reduced sperm count, impaired motility, and abnormal morphology are characteristics of oligoasthenoteratozoospermia (OAT). Genetic factors play an important role in spermatogenic failure, but the molecular basis of OAT is still not fully understood. Whole-genome sequencing (WGS) was performed in 17 infertile men with normal karyotypes and no Y-chromosome microdeletions, including 10 patients with azoospermia and 7 with OAT. Among the OAT group, three patients carrying prioritized rare variants with potential relevance to spermatogenesis were selected for detailed analysis and presentation in the current study. Variants were called using DeepVariant, and they were annotated by Ensembl Variant Effect Predictor (VEP). Variants were prioritized based on low allele frequency (< 1% in gnomAD), predicted functional impact, relevance to spermatogenesis, and consistency with a recessive inheritance model. Candidate variants were validated by Sanger sequencing and inspection of BAM files. Structural protein modeling and characterization of mutation-induced changes in molecular interactions were also performed using SWISS-MODEL and AlphaFold. We identified a hemizygous splice donor variant (c.*4 + 1G > T) in SSX1 and a missense variant (c.1190 A > G) in CPNE1 in one patient with OAT. SSX1 is a testis-expressed gene, while CPNE1 encodes a calcium-dependent phospholipid-binding protein. In another patient, a stop-gained variant (c.790 C > T) in SPTBN5 was identified, which encodes a spectrin beta chain protein. This study identifies rare candidate variants in SSX1, CPNE1, and SPTBN5 that may potentially be associated with the molecular basis of OAT. These findings provide preliminary evidence supporting further investigation of these genes in OAT and support the application of WGS in unresolved male infertility.</p>","PeriodicalId":14891,"journal":{"name":"Journal of Applied Genetics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2026-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148470789","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comprehensive genome-wide analysis of DUF668 gene family in potato reveals roles in growth and stress responses.","authors":"Aiana Gill, Tanvi Mongia, Garima Kakkar, Santosh Kumar Upadhyay, Kashmir Singh","doi":"10.1007/s13353-026-01098-z","DOIUrl":"https://doi.org/10.1007/s13353-026-01098-z","url":null,"abstract":"<p><p>The DUF668 gene family, distinguished by a plant-specific conserved domain, is significantly involved in growth and stress responses. An extensive genome-wide study of the DUF668 gene family in potato (Solanum tuberosum L.) was conducted. A total of 11 DUF668s (StDUF668-1 to StDUF668-11) were recognized and analysed for their physicochemical characteristics, chromosomal distribution, gene structures, conserved motif profiles, secondary and tertiary structures and evolutionary relationships. The StDUF668s exhibited diverse structures, with intron-exon variations contributing to their functional diversity. The analysis of promoter sequences revealed multiple cis-acting regulatory elements related to phytohormones, stress responses, and developmental processes, indicating their potential modulatory roles. Expression profiling was conducted across various potato tissues and under different hormonal and stress conditions which demonstrated that StDUF668s play a role in crucial biological processes, involving abiotic and biotic stress responses. Furthermore, qRT-PCR-based expression profiling in response to salt and osmotic stress conditions validated that StDUF668s are involved in abiotic stress responses. Protein-protein interaction and miRNA targeting analyses highlighted the molecular pathways associated with the functions of DUF668s in potato. This study advances our understanding of the DUF668 gene family and provides a valuable resource for future research aimed at improving potato stress tolerance and growth.</p>","PeriodicalId":14891,"journal":{"name":"Journal of Applied Genetics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2026-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148436723","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MGIDI-guided multi-trait selection and diversity analysis of elite and minicore-derived blackgram genotypes in the North-Western Himalayas.","authors":"Pramod Kumar Meghwal, Uttam Chandel, Nimit Kumar, Vivek Singh, Shorya Kapoor","doi":"10.1007/s13353-026-01097-0","DOIUrl":"https://doi.org/10.1007/s13353-026-01097-0","url":null,"abstract":"<p><p>Black gram is among the most ancient and important legume crop of India, yet productivity remains low due to poor harvest index, lack of genetic variability, its susceptibility to biotic and abiotic stresses. Hence, this study aimed to assess 203 elite black gram genotypes for variability, trait associations, diversity and select superior genotypes based on ideotype using multi-trait genotype-ideotype distance index. Analysis of variance revealed significant differences for all traits studied. High phenotypic and genotypic variation, coupled with high heritability and genetic advance, were observed for pods per plant, biological yield per plant and seed yield per plant, indicating substantial scope of genetic improvement. Correlation and path coefficients analyses identified biological yield per plant and pods per plant as the most important traits associated with seed yield improvement. The MGIDI selected 30 superior genotypes at 15% selection intensity and predicted favourable gains for key yield components, including seeds per pod, 100-seed weight, pods bearing branches per plant, biological yield per plant and pods per plant. Integration of MGIDI, diversity analyses and disease screening identified genotypes G108, G2, G98, G196, G107, G190, G79 and G140 as promising parental lines for breeding programmes. As the study was conducted at a single location and season, the identified genotypes require further multi-environment validation before wider deployment.</p>","PeriodicalId":14891,"journal":{"name":"Journal of Applied Genetics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2026-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148421306","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine learning integration of tissue-specific metagenomic signatures for colorectal cancer diagnosis.","authors":"Anıl Delik, Yakup Ülger, Ferhat Albayrak, Umut Orhan, Ulku Unal, Esra Gov, Sadık Dinçer","doi":"10.1007/s13353-026-01088-1","DOIUrl":"https://doi.org/10.1007/s13353-026-01088-1","url":null,"abstract":"<p><p>Colorectal cancer (CRC) represents a significant global health burden. Leveraging machine learning (ML) with metagenomic and tissue-specific data presents new opportunities for improving diagnostic accuracy and understanding the microbiome's role in CRC. This study was conducted to enhance diagnostic efficiency and identify crucial bacterial biomarkers in CRC using various ML models applied to metagenomic data. A total of 33 samples were analyzed, comprising 20 healthy controls and 13 CRC patients. Each sample included demographic data (age, gender) and bacterial information (Bacteroides, Enterococcus, Faecalibacterium, Proteobacteria, Gammaproteobacteria, Firmicutes, Enterobacteriaceae, Clostridia). Six models: Logistic Regression, Naive Bayes, Decision Tree, Support Vector Machine (SVM) with both linear and polynomial kernels and Multilayer Perceptron (MLP) were employed. Performance was evaluated using leave-one-out cross-validation (LOOCV). To address the class imbalance, F1-score was utilized as the primary metric for feature selection. A consensus-based feature elimination strategy, where bacterial features were iteratively removed only if their exclusion improved or maintained the F1-score across the majority of the models was implemented. For the MLP, a grid search was integrated into each iteration to optimize hidden layer architectures and solvers, thereby ensuring that robust performance was achieved for each feature subset. The analysis was conducted using a 10-feature initial set consisting of 2 demographic and 8 microbial features. Model performances were optimized through a consensus-based feature elimination strategy, and it was determined that diagnostic success increased with the exclusion of the Faecalibacterium, Age, and Enterobacteriaceae features during the process. The highest performance was achieved with the SVM model with Linear kernel when Bacteroides was excluded from the 9-feature subset (Table 4), reaching an accuracy of 87.88% and an F1-score of 83.33%. Within the final biomarker set, Enterococcus and Firmicutes were identified as the most critical predictive features due to the sharpest declines in F1-score observed in their absence. This study demonstrates that the systematic elimination of initial clinical and metagenomic features maximizes CRC diagnostic accuracy and model stability. The process, initiated with a 10-feature baseline set was subsequently refined to establish a high-precision diagnostic mechanism with an F1-score of 83.33%. The identified final microbial signatures, consisting of 5-6 taxa, provide a clinically applicable, non-invasive diagnostic foundation with low input requirements.</p>","PeriodicalId":14891,"journal":{"name":"Journal of Applied Genetics","volume":" ","pages":""},"PeriodicalIF":2.1,"publicationDate":"2026-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148421335","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}