>program name
oligo/dyad-analyis
>data set
yst01
>parameters
-v 1
-thosig 0
-purge
-org Saccharomyces_cerevisiae
-seq Saccharomyces_cerevisiae_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/rand_gene_selections
>postprocessing
>data set
yst02
>parameters
-v 1
-thosig 0
-purge
-org Saccharomyces_cerevisiae
-seq Saccharomyces_cerevisiae_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/rand_gene_selections
>postprocessing
>data set
yst03
>parameters
-v 1
-thosig 0
-purge
-org Saccharomyces_cerevisiae
-seq Saccharomyces_cerevisiae_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/rand_gene_selections
>postprocessing
We selected the words assembled in the most significant pattern
cluster (GAGTCA, GAGTCAA, AGTCAT), and set the threshold to 0.74,
which is the largest pattern of this cluster (heptanucleotide), and
encompasses the two other ones. This threshold was selected to
filter out AGTCAT, which probably reflects the flanking preferences
of the motif, but does not contain the most significant core
(GAGTCA).

>data set
yst04
>parameters
-v 1
-thosig 0
-purge
-org Saccharomyces_cerevisiae
-seq Saccharomyces_cerevisiae_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/rand_gene_selections
>postprocessing
We selected the most significant pattern
>data set
yst05
>parameters
-v 1
-thosig 0
-purge
-org Saccharomyces_cerevisiae
-seq Saccharomyces_cerevisiae_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/rand_gene_selections
>postprocessing
If we have to select one, it's the octanucleotide cccacaga. Its sig is low, but give a higher sig with the
binomial model (0.81), and there are few sequences (3)
>data set
yst06
>parameters
-v 1
-thosig 0
-purge
-org Saccharomyces_cerevisiae
-seq Saccharomyces_cerevisiae_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/rand_gene_selections
>postprocessing
The three most significant patterns assemble in all cases but one in 1 nonanucleotide.
One occurence of attagga is isolated, and has been retained because the pattern is quite big (7 nt) and could be
the core of the binding site
cluster (aattagga, attagga, attaggaa).
>data set
yst07
>parameters
-v 1
-thosig 0
-purge
-org Saccharomyces_cerevisiae
-seq Saccharomyces_cerevisiae_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/rand_gene_selections
>postprocessing
>data set
yst08
>parameters
-v 1
-thosig 0
-purge
-org Saccharomyces_cerevisiae
-seq Saccharomyces_cerevisiae_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/rand_gene_selections
>postprocessing
In this case, we found 2 clusters which are susceptible to be regulatory sequences :
In the first one, the hexanucleotide aggaag seem not to be sufficient : There are a lot of
occurences in some sequences. On the contrary, the heptanucleotide taggaag can be found in 7/11 sequences,
with 2 occurences in 1 sequence only.
In the second, the octanucleotide gccagaaa seem to be a good candidate (big motif,1 occurence in 7/11 sequences)
the other words which are part of this cluster are discarded because they are essentially polyA, and have only a small
part overlapping with the octanucleotide
>data set
yst09
>parameters
-v 1
-thosig 0
-purge
-org Saccharomyces_cerevisiae
-seq Saccharomyces_cerevisiae_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/rand_gene_selections
>postprocessing
Nothing convincing
>data set
yst10
>parameters
-v 1
-thosig 0
-purge
-org Saccharomyces_cerevisiae
-seq Saccharomyces_cerevisiae_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Saccharomyces_cerevisiae/rand_gene_selections
>postprocessing
The only pattern with an acceptable sig is atatag, but it's an hexanucleotide, and its distribution
amongst the sequences (6 occurences in one sequence) seems bad.
>data set
dm01
>parameters
-v 1
-thosig 0
-purge
-org Drosophila_melanogaster
-seq Drosophila_melanogaster_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/rand_gene_selections
>postprocessing
For the most significant cluster of words, we only selected the
heptanucleotide ACAGCAA, because hexanucleotides are likely to return false
positions.  We also choose the 2 octanucleotides CACGGCCA and CGCTTTAC
which have a reasonably good significance. 
>data set
dm02
>parameters
-v 1
-thosig 0
-purge
-org Drosophila_melanogaster
-seq Drosophila_melanogaster_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/rand_gene_selections
>postprocessing
This is a very delicate case because there is only one sequence. The
three oligonucletides have a quite low sig There are however 2 dyads
cggn{8}ggc and cggn{9}gcg which have a high sig (3.03 and 2.04), and
which can be assembled to form CGGn{8}GGCG. This site seems
reasonable (3 occurences).
>data set
dm03
>parameters
-v 1
-thosig 0
-purge
-org Drosophila_melanogaster
-seq Drosophila_melanogaster_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/rand_gene_selections
>postprocessing
>data set
dm04
>parameters
-v 1
-thosig 0
-purge
-org Drosophila_melanogaster
-seq Drosophila_melanogaster_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/rand_gene_selections
>postprocessing

>data set
dm05
>parameters
-v 1
-thosig 0
-purge
-org Drosophila_melanogaster
-seq Drosophila_melanogaster_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/rand_gene_selections
>postprocessing
We selected the 2 most significant patterns, because they has a
reasonable repartition (occurences in the 3 sequences).

>data set
dm06
>parameters
-v 1
-thosig 0
-purge
-org Drosophila_melanogaster
-seq Drosophila_melanogaster_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/rand_gene_selections
>postprocessing
>data set
dm07
>parameters
-v 1
-thosig 0
-purge
-org Drosophila_melanogaster
-seq Drosophila_melanogaster_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/rand_gene_selections
>postprocessing
The most significant pattern was chosen; its distribution seems reasonable.
>data set
dm08
>parameters
-v 1
-thosig 0
-purge
-org Drosophila_melanogaster
-seq Drosophila_melanogaster_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Drosophila_melanogaster/rand_gene_selections
>postprocessing
After observation of the maps with pentanucleotides and
heptanucleotides with sig>=-1, it seems reasonable to think that the
hexanucleotide tagtaa is the constant core of a binding site, but is
not sufficient (many occurences - 6 in the first sequence, 2 in the
second, and 5 in the third). We selected the 3 heptanucleotides
forming the assembly made of 5nt + 6nt.

>data set
mus01
>parameters
-v 1
-thosig 0
-purge
-org Mus_musculus
-seq Mus_musculus_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/rand_gene_selections
>postprocessing
The only significant pattern is not convincing (4 occurences in 1
sequence and 0 in the 2 other ones).
>data set
mus02
>parameters
-v 1
-thosig 0
-purge
-org Mus_musculus
-seq Mus_musculus_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/rand_gene_selections
>postprocessing
In most mouse families, there is not a single significant pattern,
when the threshold of sig >=0 is selected. This suggests that our
calibration might be too stringent. We lowered the threshold to -1
to evaluate slightly less significant patterns. With mus02, we
selected the most significant pattern, which is detected as
heptanucleotide TCATGCA (sig=-0.37) and as hexanucleotide (CATGCA).
>data set
mus03
>parameters
-v 1
-thosig 0
-purge
-org Mus_musculus
-seq Mus_musculus_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/rand_gene_selections
>postprocessing
In most mouse families, there is not a single significant pattern,
when the threshold of sig >=0 is selected. This suggests that our
calibration might be too stringent. We lowered the threshold to -1
to evaluate slightly less significant patterns. 
In mus03, a single heptanucleotide had a sig >= -1, we thus retained
it.
>data set
mus04
>parameters
-v 1
-thosig 0
-purge
-org Mus_musculus
-seq Mus_musculus_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/rand_gene_selections
>postprocessing
>data set
mus05
>parameters
-v 1
-thosig 0
-purge
-org Mus_musculus
-seq Mus_musculus_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/rand_gene_selections
>postprocessing
TO DISCUSS
>data set
mus06
>parameters
-v 1
-thosig 0
-purge
-org Mus_musculus
-seq Mus_musculus_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/rand_gene_selections
>postprocessing
>data set
mus07
>parameters
-v 1
-thosig 0
-purge
-org Mus_musculus
-seq Mus_musculus_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/rand_gene_selections
>postprocessing
In most mouse families, there is not a single significant pattern,
when the threshold of sig >=0 is selected. This suggests that our
calibration might be too stringent. We lowered the threshold to -1
to evaluate slightly less significant patterns. In mus07, we
selected the two most significant heptanucleotides: GAGGCCC and
CTCGCTC. Hexanucleotides were not retained because their
significance is weaker, and selecting them would result in many
matches, likely to be spurious.

>data set
mus08
>parameters
-v 1
-thosig 0
-purge
-org Mus_musculus
-seq Mus_musculus_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/rand_gene_selections
>postprocessing
>data set
mus09
>parameters
-v 1
-thosig 0
-purge
-org Mus_musculus
-seq Mus_musculus_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/rand_gene_selections
>postprocessing
>data set
mus10
>parameters
-v 1
-thosig 0
-purge
-org Mus_musculus
-seq Mus_musculus_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/rand_gene_selections
>postprocessing
This pattern has a low sig, but is the only one with sig > -1, and
has its occurences have good repartition over the input sequences.
>data set
mus11
>parameters
-v 1
-thosig 0
-purge
-org Mus_musculus
-seq Mus_musculus_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/rand_gene_selections
>postprocessing
>data set
mus12
>parameters
-v 1
-thosig 0
-purge
-org Mus_musculus
-seq Mus_musculus_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Mus_musculus/rand_gene_selections
>postprocessing
>data set
hm01
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
>data set
hm02
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
The only significant pattern, CAGCACAC, is weakly significant
(sig=0.52) and is only found in 3 out of 9 sequences. 
>data set
hm03
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
The only significant pattern, ATAGTCTG, is weakly significant
(sig=0.29), and it is found in 3/10 sequences only.
>data set
hm04
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
-minol 8 -maxol 8
>postprocessing
The analysis of 5nt, 6nt and 7nt detected dozens of patterns
covering the whole sequence, suggesting that the calibration is
inappropriate for this data set (same as for hm18). We restricted
the analysis to octanucleotides, and selected the most significant
pattern assembly made of two octanucleotides AAACCCGC (1.52) and
AACCCGCA (1.68). The octanucleotide GGCGGAGA was also retained
because it is reasonably significnt, and found in most sequences.
>data set
hm05
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
No significant pattern is detected with sig >=0, probably dur to the
small number of sequences. When the threshold is lowered to 0, some
patterns are selected. The most significant among these, TAGGAGT, is
found in the three sequences, and in 2/4 cases, it is extended by an
overlapping pattern GGAGTAT. This could be a reasonable candidate.
>data set
hm06
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
>data set
hm07
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
A single heptanucleotide is reasonably significant, and is found in 4/5 sequences. 
>data set
hm08
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
We only retained the most significant pattern (TGACGT). 
>data set
hm09
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
>data set
hm10
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
The selected patterns had a very weak significance, and each was
found on a small number of sequences only.
>data set
hm11
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
>data set
hm12
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
Two octanucleotides (AATTGACG and ACGTCATG) were selected with
sig>=0, but present only on one sequence (seq_0). To improve
sensitivity, we looked at patterns with sig >= -1. Many 6nt, 7nt and
8nt were selected with sig >= -1, which aligned with the core
octanucleotide AATTGACG. We selected the oligos of this cluster for
pattern matching, which allowed to predict a match on seq_1, in
addition to the two matche found in seq_0 with sig >= 0.
>data set
hm13
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
>data set
hm14
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
Not really convincing, but it's not easy to find overrepresented
motifs in a set of 2 sequences only.  We think this octanucleotide
could be a TF binding site because of the distribution of its
occurences.
>data set
hm15
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
>data set
hm16
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
>data set
hm17
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
All patterns detected align with the cor AATTCCCG (sig= 3.56), and
their matching locations overlap. We retain all of them. 
>data set
hm18
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
-minol 7 -maxol 8
>postprocessing
For this data set, the analysis of 5nt and 6nt returned too many
patterns, which covered the whole sequence. When the analysis is
resticted to 8nt only, a single pattern (GCGGAGAA, sig=1.83) is
selected, which covers 3 sequences out of 5. When 7nt and 8nt are
taken together, the second most significant pattern (ACTGCGC,
sig=1.32) is assembled with several other significant patterns. We
retained this pattern cluster centered on ACTGCGC, and the
significant octanucleotide GCGGAGAA.
>data set
hm19
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
>data set
hm20
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
>data set
hm21
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
>data set
hm22
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
The 2 most significant patterns could be relevant
>data set
hm23
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
Not really convincing, we simply selected the best match
>data set
hm24
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
The selected patterns had a very weak significance, and each was
found on a small number of sequences only.
>data set
hm25
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing

The only pattern with significance > 0 is not convincing (5
occurences in 1 sequence and 0 in the other one). We then chose the
patterns gctata and aaggac with lower sig, but with a good
repartition (1 occurence in seq 0, and 2 in seq 1, with similar
distance from the start codon)
>data set
hm26
>parameters
-v 1
-thosig 0
-purge
-org Homo_sapiens
-seq Homo_sapiens_files.txt
-outdir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/multi/calibN_bg-purge
-2str
-minol 6
-maxol 8
-minsp 0
-maxsp 20
-bg calibN
-sort score
-task report,synthesis
-noov
-calib_dir /Users/jvanheld/motif_discovery_competition_2003/results/Homo_sapiens/rand_gene_selections
>postprocessing
We selected the two most significant patterns, which have a sig > 1.