Table 1.
ChIPtig statistics and results of motif finding in the mouse ESC dataset using Velvet
| TF | No. ChIPtigs | Mapped ChIPtigs (%) | Motif rank with peak-calling | Motif rank with de novo pipeline | Motif rank with random ChIPtigs |
|---|---|---|---|---|---|
| c-MYC | 5,159 | 92.9 | 1 | 1 | 4 |
| CTCF | 2,152 | 92.9 | 1 | 2 | 1 |
| ESRRB | 30,278 | 95.9 | 1 | 1 | 1 |
| KLF4 | 1,660 | 90.4 | 1 | 1 | 1 |
| NANOG | 5,163 | 94.3 | 1 | N | N |
| n-MYC | 3,610 | 86.8 | 1 | 1 | 1 |
| POU5F1 | 2,528 | 92.9 | 1 | 1 | N |
| SMAD1 | 596 | 96.1 | 7 | N | N |
| SOX2 | 2,511 | 92.5 | 1 | 1 | N |
| STAT3 | 4,329 | 94.8 | 1 | 1 | N |
| TCFCP2I1 | 20,566 | 95.7 | N | N | N |
| ZFX | 3,348 | 93.5 | 1 | 1 | 1 |
Three settings were evaluated for motif finding performance: peak-calling using reference genome (MACS), top 1,000 ChIPtigs from the de novo pipeline, and 1,000 random ChIPtigs from the same experiment assembled by Velvet. The rank of the known motif (from JASPAR) in the DREME results is shown for each TF. ‘N’ in a row means that either DREME did not find any motif, or none of the motifs found by DREME matches the known motif for the TF in that row