RESUMO
Current gold standard diagnostic strategies are unable to accurately differentiate malignant from benign small renal masses preoperatively; consequently, 20% of patients undergo unnecessary surgery. Devising a more confident presurgical diagnosis is key to improving treatment decision-making. We therefore developed MethylBoostER, a machine learning model leveraging DNA methylation data from 1228 tissue samples, to classify pathological subtypes of renal tumors (benign oncocytoma, clear cell, papillary, and chromophobe RCC) and normal kidney. The prediction accuracy in the testing set was 0.960, with class-wise ROC AUCs >0.988 for all classes. External validation was performed on >500 samples from four independent datasets, achieving AUCs >0.89 for all classes and average accuracies of 0.824, 0.703, 0.875, and 0.894 for the four datasets. Furthermore, consistent classification of multiregion samples (N = 185) from the same patient demonstrates that methylation heterogeneity does not limit model applicability. Following further clinical studies, MethylBoostER could facilitate a more confident presurgical diagnosis to guide treatment decision-making in the future.
RESUMO
Gene expression is an important phenotype that informs about genetic and environmental effects on cellular state. Many studies have previously identified genetic variants for gene expression phenotypes using custom and commercially available microarrays. Second generation sequencing technologies are now providing unprecedented access to the fine structure of the transcriptome. We have sequenced the mRNA fraction of the transcriptome in 60 extended HapMap individuals of European descent and have combined these data with genetic variants from the HapMap3 project. We have quantified exon abundance based on read depth and have also developed methods to quantify whole transcript abundance. We have found that approximately 10 million reads of sequencing can provide access to the same dynamic range as arrays with better quantification of alternative and highly abundant transcripts. Correlation with SNPs (small nucleotide polymorphisms) leads to a larger discovery of eQTLs (expression quantitative trait loci) than with arrays. We also detect a substantial number of variants that influence the structure of mature transcripts indicating variants responsible for alternative splicing. Finally, measures of allele-specific expression allowed the identification of rare eQTLs and allelic differences in transcript structure. This analysis shows that high throughput sequencing technologies reveal new properties of genetic effects on the transcriptome and allow the exploration of genetic effects in cellular processes.