Colorectal cancer stands as a leading cause of cancer-related morbidity and mortality globally,with liver metastases being a significant determinant of patient prognosis.Conventional diagnostic methods,includ-ing imaging studies and biomarker testing,frequently exhibit inadequate sensitivity and specificity,underscoring the necessity for more advanced technologies.Recent advancements in genomics,transcriptomics,proteomics,me-tabolomics,and epigenomics have revolutionized our understanding of the biological mechanisms driving colorectal cancer.These methodologies enable comprehensive analyses of genetic mutations,gene expression profiles,protein modifications,and metabolic reprogramming,all of which are pivotal to the metastatic process.This article high-lights the advanced capabilities of artificial intelligence(AI)technologies in processing complex multi-omics data,thereby enhancing diagnostic accuracy and supporting personalized treatment strategies.It also addresses the challenges AI encounters in multi-omics analyses,such as ensuring data quality,improving model interpretability,and facilitating clinical translation.Additionally,it explores the potential integration of emerging technologies like single-cell sequencing and spatial omics into large-scale,multicenter studies to further enhance the clinical utility of these tools.

Colorectal cancer stands as a leading cause of cancer-related morbidity and mortality globally,with liver metastases being a significant determinant of patient prognosis.Conventional diagnostic methods,includ-ing imaging studies and biomarker testing,frequently exhibit inadequate sensitivity and specificity,underscoring the necessity for more advanced technologies.Recent advancements in genomics,transcriptomics,proteomics,me-tabolomics,and epigenomics have revolutionized our understanding of the biological mechanisms driving colorectal cancer.These methodologies enable comprehensive analyses of genetic mutations,gene expression profiles,protein modifications,and metabolic reprogramming,all of which are pivotal to the metastatic process.This article high-lights the advanced capabilities of artificial intelligence(AI)technologies in processing complex multi-omics data,thereby enhancing diagnostic accuracy and supporting personalized treatment strategies.It also addresses the challenges AI encounters in multi-omics analyses,such as ensuring data quality,improving model interpretability,and facilitating clinical translation.Additionally,it explores the potential integration of emerging technologies like single-cell sequencing and spatial omics into large-scale,multicenter studies to further enhance the clinical utility of these tools.

Objective To study the distributive characteristics of HLA-A,B,DRBI alleles and haplotypes patients with ALL in southern Chinese Han.Methods The frequencies of HLA-A,B,DRB1alleles and haplotypes were estimated by Expectation-Maximization method based on the genotypes of 572patients with ALL and 5645 unrelated health donors,and then compared by chi-square test.Results The frequencies of HLA-A33(7.15%vs 9.3%,OR=0.73,P＜0.05),B58(5.93%vs 8.75%,OR=0.64,P＜0.05),DRB1*17(5.15%vs 6.30%,OR=0.82,P＜0.05)alleles and HLA-A33-B58-DRB1*17(2.46%vs 4.14%,OB=0.35,P＜0.05)haplotype were significantly lower in ALL patient groups than that in controls.The frequencies of HLA-A3(2.1%vs 1.26%,OR=1.7,P＜0.05),B51(7.25%vs 5.78%,OR=1.3,P＜0.05)and DRB*12 (16.13%vs 12.99%,OR=1.35,P＜0.05)alleles and A2-B51-DRB1*12(1.24%vs.0.89%,OR=1.66,P＜0.05)haplotype were significantly higher in ALL patient groups than that in controls.Conclusion These results indieated that HLA-A33-B58-DRB1*17 haplotype was a associated with a diminished incidence of ALL.and HLA-A3 auele or A2-B51-DRB1*12 haplotype was weakly associated with ALL.