Skip to main content
Fig. 3 | Genome Medicine

Fig. 3

From: Prioritization and functional assessment of noncoding variants associated with complex diseases

Fig. 3

Evaluation of features used in PAFA. a A Venn diagram on variant sets from GWASdb and 1000 Genomes. The overlapping section represents the mutations that exist in both GWASdb and 1000 Genomes. b FST scores for cSNPs in GWASdb versus variants in 1000 Genomes after removing variants in GWASdb. c Tenfold cross-validations are applied to evaluate the performance of PAFA and three PAFA classifiers that were separately constructed on the three feature groups. Receiver operating characteristic (ROC) curves are exhibited. The dashed line indicates random chance. The value of the area under the curve (AUC) is calculated for each feature. d Violin plots of PAFA scores for three disease-related variant sets, including ovarian serous cystadenocarcinoma (OSC), age-related macular degeneration (AMD), and type 2 diabetes (T2D). PAFA scores are provided by the classifier that was constructed based on all features (PAFA), all except allele frequencies (-AF) and all except FST and dispersion score (-FST/DS). The p values for comparisons of these scores were calculated by Wilcoxon rank-sum test

Back to article page