. 2016 Dec 20;111(12):2551–2561. doi: 10.1016/j.bpj.2016.11.014

Table 3.

Comparison between Pulldown Efficiencies for Each Base at Positions Before the mCpG and the Corresponding Pulldown Efficiency at Position −20

Position	A	T	C	G
−20	0	0	0	0
−19	0.006140	−0.004816	−0.008181	0.005322
−18	0.001259	−0.001074	0.003418	−0.003020
−17	0.010141	−0.003417	−0.005439	−0.005147
−16	0.007004	0.002060	−0.004954	−0.009631
−15	0.007626	−0.003615	−0.003336	−0.002128
−14	0.002729	−0.003517	0.000819	0.002769
−13	0.009267	−0.001765	−0.008163	−0.004113
−12	0.001399	0.003566	0.000795	−0.009399
−11	0.009058	−0.001970	−0.010484	−0.002031
−10	0.003779	−0.000941	0.002484	−0.008548
−9	0.005571	0.001141	−0.000707	−0.011832
−8	0.005654	−0.003704	0.001875	−0.004070
−7	0.010948	−0.000573	−0.006428	−0.012071
−6	0.011101	0.003206	−0.006664	−0.014880
−5	0.013346	−0.007117	−0.001714	−0.006418
−4	0.001895	0.004467	−0.002076	0.000427
−3	0.023423^a	0.013456	0.002463	0.020995^a
−2	0.018563^a	0.025926^a	−0.022233^a	0.018484
−1	0.019887^a	−0.053899^a	0.021795^a	0.024648^a

Shown is the deviation from unity in the ratio $E_{i} (b) / E_{- 20} (b)$ for base b at position i labeled with respect to the mCpG. Values are not considered statistically significant, unless otherwise indicated. Fig. 6 visualizes this data with statistically significant values outlined in black.

Considered statistically significant, with p value < 0.000125.