Figure 6
From: Assessing the impact of human genome annotation choice on RNA-seq expression estimates
Statistics for comparing RNA-seq-based fold-changes and RT-qPCR-based fold-changes vary with different annotations. The comparison of fold-change estimates between RNA-seq and RT-qPCR using two RNA-seq expression estimates and three statistics. (a) TPM estimates are produced by the OSA package. (b) FPKM estimates are generated by Cufflinks with TopHat alignment. The RefSeq annotation always has the lowest (1) average absolute deviation and (2) root mean square error, and the highest correlation coefficient when treating RT-qPCR estimates as the ground truth.