Biclustering of Expression Data
Yizong Cheng and George M. Church
Supplementary Information
- Yeast expression matrix, based on
Tavazoie et al. 8224 rows and 17 columns, 4 bytes for each element,
with -1 indicating a missing value.
- Human expression matrix, based on
Alizadeh et al.
with 4026 rows and 96 columns, 4 bytes for each element, and 999
indicating a missing value.
- 100 biclusters from the yeast data.
Each bicluster is documented with three rows. The first row contains
the number of rows, the number of columns, and the score. The second row
is the column set and the third the row set. If a row index in the
row set is preceded by a minus sign, the additive inverse was used
in the process. All indices start at 0.
- 100 biclusters from the human data
- Genes associated with rows of the yeast data
- Comparison of the yeast biclusters and clusters in
Tavazoie et al.,
After the index of each of the 100 biclusters, 30 numbers indicate the
frequencies at which the elements occur in the 30 clusters of Tavazoie et al.
- biclulym biclustering program executable on Solaris 7 for the
human data. The user has to provide data filename and two thresholds (alpha
in Algorithm 1 and delta in Algorithm 2) as command line arguments.
To get the 100 biclusters listed above, use thresholds 1.2 and 1200.
This version only works for the human data. Note: the program be renamed
depending on your browser!
- bcplot plotting program executable on Solaris 7 for the
output of the previous program. Note: the program be renamed
depending on your browser!