Columns in weighted mappings files
   1. uniq_tag_sequence - unique CAGE tag sequences
   2. tag_count_library - total tag counts in library. SUM( tag_count_timecourse ) = tag_count_library
   3. edit_string - what type of editing was performed on tag sequence when mapping to chromosome
   4. chr - chromosome
   5. strand - tag direction
   6. start - starting position
   7. end - ending position
   8. percentage - percent match (always 100)
   9. map_pos - number of locations this tag mapped to the chromosome
  10. ribo_flag - tag mapped to ribosome equal or better than chromosome?
  11. refseq_flag - tag mapped to refseq better than chromosome?
  12. time_course - time course which tag appeared
  13. tag_count_timecourse - number of tags with specified timecourse
  14. tpm_in_ribo - tag per million per timecourse including ribosome flagged tags
  15. tpm_ex_ribo - tag per million per timecourse excluding ribosome_flagged tags
  16. weight - weight of expression
  17. rescue_weight - weight from "guilt-by-association" rescue

i.e. to calculate total mapped tags ignore ribo_flag = RIBOSOME tags and multiply column 13 by column 17 (count by weight). To calculate signal (tpm) for a tag in a position multiply column 15 by column 17.


Columns in normalised tag clusters files:

1. chromosome
2. start
3. stop
4. strand

For human:
5. to 16. tpm values
17. to 28. counts

For mouse:
5. to 17. tpm values
18. to 30. counts

Columns are in same order as Fig 1a and Fig 1b. For all mapping and clustering methods see Methods section.