Transcription initiation peaks based on FANTOM5 CAGE data on mm10

== Overview

Decomposition-based peak identification (DPI, https://github.com/hkawaji/dpi1) is applied to the re-processed (re-aligned) FANTOM5 data, upon mm10 (GRCm38), obtained from below:

http://fantom.gsc.riken.jp/5/datafiles/reprocessed/mm10_v1/basic/
The same parameters to the ones used in the previous paper (Forrest ARR, Kawaji H, Rehli M, et al. Nature 507: 462–470, 2014) was used.

== Data files

Four data files per assembly are prepared as below.

tag cluster in the original definition (*.tc.bed.gz)
full set of DPI peaks (*.tc.decompose_smoothing_merged.bed.gz)
permissive set of DPI peaks (*.tc.decompose_smoothing_merged.ctssMaxCounts3.bed.gz)
robust set of DPI peaks (*.tc.decompose_smoothing_merged.ctssMaxCounts11_ctssMaxTpm1.bed.gz)