Skip to main content

Table 2 Updating linker length improves prediction - Gamma linker length model

From: Predicting nucleosome positioning using a duration Hidden Markov Model

   1st order     4th order  
update total sensitivity(%) FDR(%) update total sensitivity (%) FDR(%)
0 8670 (22) 59 (0.7) 32 (0.7) 0 8896 (27) 64 (0.5) 28 (0.5)
1 9347 (25) 65 (0.8) 30 (0.7) 1 9550 (42) 70 (0.6) 26 (0.6)
4 9833 (39) 67 (0.7) 31 (0.7) 4 9880 (35) 72 (0.5) 27 (0.6)
  1. Total predictions, sensitivity, and false discovery rate (FDR) are the averages (standard deviations in parentheses) based on 10 repeated simulations. For each simulation a genomic sequence consisting of 10000 nucleosomes and 10001 linkers were simulated using the 1st and 4th order yeast models. The linker length distribution was initialized as uniform on 1,..., 200, and was iteratively updated in the dHMM. Results are shown after 0, 1, and 4 updates.