Closed bschilder closed 3 years ago
Take our CUT&TAG sample that (accidentally) had 146 million reads. Computationally downsample the reads and repeat the Precision-Recall curves for these different read depths.
You can also plot the F1 score (merged Precision-Recall score), as in Fig 3b of the SEACR paper.
Relatedly, you will want to see how downsampling affects the duplication rate.
Take our CUT&TAG sample that (accidentally) had 146 million reads. Computationally downsample the reads and repeat the Precision-Recall curves for these different read depths.
You can also plot the F1 score (merged Precision-Recall score), as in Fig 3b of the SEACR paper.
Relatedly, you will want to see how downsampling affects the duplication rate.