This directory contains information about whether Fantom and public
cDNA sequences represent full-length RNAs or artefactual truncations.

f3_mm5_full-length : 41025 out of 102801 Fantom cDNAs that have
experimental support for their 5' and 3' ends (see below), and are not
immediately upstream of A-rich sequences in the genome (> 10 As within
20 nt).

gb_mm5_full-length : 26818 out of 56006 public non-RIKEN cDNAs that
have experimental support for their 5' and 3' ends (see below), and
are not immediately upstream of A-rich sequences in the genome (> 10
As within 20 nt).

f3_mm5_tails : Genomic sequences from -10 to +30 relative to the 3'
ends of Fantom1+2+3 sequences. When they are A-rich, it suggests that
the Fantom sequence may be derived from internal priming of genomic
DNA, pre-mRNA, or mature RNA.

gb_mm5_tails : Genomic sequences from -10 to +30 relative to the 3'
ends of public non-RIKEN cDNAs. Note that non-RIKEN cDNAs often
include a poly-A tail, but Fantom cDNAs rarely do.

f3_mm5_5end_support.txt :
f3_mm5_3end_support.txt :
gb_mm5_5end_support.txt :
gb_mm5_3end_support.txt :
Presence or absence of various kinds of experimental support for the
5' and 3' ends of the cDNAs. cDNAs were not considered unless they
mapped unambiguously to the genome, and had at most 5 nt of sequence
unaligned at the end being considered (5' or 3'), not counting poly-A
tails for non-RIKEN sequences.

Criteria for 5' ends:
2 CAGE tag starts within +-15 nt
3 CAGE tag starts within +-60 nt
4 CAGE tag starts within +-100 nt
1 GSC ditag start within +-0 nt
2 GSC ditag starts within +-50 nt
1 GIS ditag start within +-15 nt
1 RIKEN 5'EST start within +-3 nt
2 RIKEN 5'EST starts within +- 100 nt
1 non-RIKEN 5'EST start within +-2 nt
2 non-RIKEN 5'EST starts within +-100 nt
1 other Fantom clone start within +- 25 nt
1 non-RIKEN RNA start within +-50 nt

Criteria for 3' ends:
1 GSC ditag end within +-0 nt
2 GSC ditag ends within +-50 nt
1 GIS ditag end within +-15 nt
1 RIKEN 3'EST end within +-2 nt
2 RIKEN 3'EST ends within +-100 nt
1 non-RIKEN 3'EST end within +-7 nt
2 non-RIKEN 3'ESTs ends within +-100 nt
1 other Fantom clone end within +- 25 nt
1 non-RIKEN RNA end within +-50 nt

Each of these criteria has less than 1 in 1000 chance of occurring if
the sites are randomly scattered across the genome.  Sequences from
the same clone (e.g. ESTs) were not counted as independent
evidence. Both orientations were considered for GSC ditags.

Many thanks to Par Engstrom for help with ESTs, and Shintaro Katayama
and Akira Hasegawa for help with tags. Contact: Martin Frith