🧬 Can fewer genetic variants provide better population-specific risk prediction? Polygenic risk scores for venous thromboembolism (VTE) have largely been developed using European ancestry datasets. Their performance in other populations may therefore be limited. A Phenomics study introduces a self-adapting PRS framework tailored to Han Chinese populations. The researchers evaluated VTE-associated SNPs and developed models incorporating either 318 variants or a selected subset of 44 variants. In an independent Chinese testing cohort: • 318-SNP model: AUC = 0.739 • 44-SNP model: AUC = 0.709 • Conventional PRS models: AUC = 0.620–0.718 The 44-SNP model retained substantial predictive performance despite using far fewer genetic features. The study also found that the self-adapting PRS provided risk-stratification information beyond selected clinical characteristics, including age and smoking status. These findings demonstrate the value of considering population-specific genetic architectures when developing risk prediction tools. Clinical utility will require further prospective evaluation and feasible screening strategies. 🔗 doi.org/10.1007/s43657-024-0… #Phenomics #Genomics #PolygenicRiskScore #VTE

🧬 Can pyroptosis-related lncRNAs help predict ovarian cancer prognosis? Pyroptosis is an inflammatory form of programmed cell death that may influence tumor biology. However, the prognostic roles of pyroptosis-related long non-coding RNAs (lncRNAs) in ovarian cancer remain incompletely understood. A study published in Phenomics developed a five-lncRNA prognostic signature using TCGA and GTEx datasets. Key findings: • The model distinguished patients with different overall and progression-free survival outcomes. • It outperformed 16 existing prognostic models in the authors' comparative analysis. • Lower-risk patients showed higher immune infiltration and homologous recombination deficiency scores. • Molecular clustering identified three ovarian cancer subtypes with distinct survival and immune characteristics. • Expression of the five lncRNAs was further examined by qRT-PCR in ovarian cancer cells and tissues. These results highlight the potential of integrating cell-death-related transcriptomic signatures with clinical phenotypes to better characterize ovarian cancer heterogeneity. Further validation will be necessary before clinical application. 🔗 doi.org/10.1007/s43657-024-0… #Phenomics #OvarianCancer #Pyroptosis #lncRNA

🦠 NetNiche integrates prior biological knowledge with graph representation learning to infer microbe–metabolite and microbe–microbe interactions, outperforming several established methods across simulated and real microbiome datasets. DOI: 10.1007/s43657-024-00168-8 #Phenomics

🧬 A Phenomics Mendelian randomization study finds genetically predicted ALS was associated with reduced colorectal cancer risk in East Asian datasets, while no causal link was found with overall cancer or most other cancers. DOI: 10.1007/s43657-024-00159-9 #Phenomics #ALS