List of test datasets

[kopardev]

1. TF narrow peaks CTCF in human

• 3 replicates IP
• 3 replicates Input
• Part of a larger time series dataset
• Found using these filters

2. TF narrow peaks JUN in human

• 3 IP replicates
• 3 Input replicates
• Found using these filters

3. H3K4me3 narrow promoter peaks in human

• 2 IP replicates
• 2 Input replicates
• Found using these filters

4. H3K27ac enhancer (mixed=narrow+broad) peaks in human

• 2 IP replicates
• 2 Input replicates (same as above.)
• Found using these filters

5. H3K36me3 broad peaks in human associated with heterochromatin

• 2 IP replicates
• 2 Input replicates (same as above.)
• Found using these filters

[kopardev]

@kelly-sovacool:

• Download these from SRA
• systematically call peaks
• use them for testing pipeline
• use them for testing RECAP + MSPC strategy

[kelly-sovacool]

For differential analysis, use dataset 1 and pick two time points. Check out the paper to compare results.

[kelly-sovacool]

For all, compare peak calls to Encode peaks (macs2)

• Conservative peaks bed files
• make a venn diagram showing overlap between peaks from ours vs encode pipeline
• compare total # of peaks

[kelly-sovacool]

• Compare peak overlaps between champagne and encode wtih Intervene
• Use MEME FIMO to determine unique peaks that contain the motif. Are we missing peaks, is encode missing peaks?