Table 1 SVM Features
From: Cell-type specificity of ChIP-predicted transcription factor binding sites
No | Name | Group | Comment |
|---|---|---|---|
1 | Height | Height | Peak height (percentiles) |
2 | Length | Length | Peak length |
3 | Promoter | Promoter | Overlap with promoter (boolean) |
4 | TSS dist | TSS dist | Dist. to closest transcription start site (max 20.000) |
5 | Cluster TFs | Cluster | Number of TFs in overlapping cluster |
6 | Cluster avg height | Cluster | Avg peak height in overlapping cluster |
7 | Chromatin avg | Chromatin | Avg DNase signal in two cell types |
8 | Chromatin diff | Chromatin | DNase signal diff between two cell types |
9 | H3K4me3 avg | H3K4me3 | Avg H3K4me3 signal in two cell types |
10 | H3K4me3 diff | H3K4me3 | H3K4me3 signal diff between two cell types |
11 | H3K27me3 avg | H3K27me3 | Avg H3K27me3 signal in two cell types |
12 | H3K27me3 diff | H3K27me3 | H3K27me3 signal diff between two cell types |
13 | CpG freq | CpG | CpG frequency in peak region sequence |
14 | High CpG | CpG | Equal to 1 if sequence is high in CpG |
15 | Low CpG | CpG | Equal to 1 if sequence is low in CpG |
16 | PhyloP | PhyloP | PhyloP conservation score in sequence region |