Mercurial > repos > rnateam > blockclust

comparison blockclust.xml @ 4:49e600128a73 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author rnateam
date Wed, 09 Jul 2014 08:38:01 -0400
parents 27dde42069e0
children 6721468f2f9f
comparison
equal deleted inserted replaced
3:27dde42069e0 4:49e600128a73
159 processing patterns. We propose a novel way to encode expression profiles 159 processing patterns. We propose a novel way to encode expression profiles
160 in compact discrete structures, which can then be processed using 160 in compact discrete structures, which can then be processed using
161 fast graph-kernel techniques. BlockClust allows both clustering and 161 fast graph-kernel techniques. BlockClust allows both clustering and
162 classification of small non-coding RNAs. 162 classification of small non-coding RNAs.
163 163
164 BlockClust runs in three modes: 164 BlockClust runs in three operating modes:
165
165 1) Pre-processing - converts given mapped reads (BAM) into BED file of tags 166 1) Pre-processing - converts given mapped reads (BAM) into BED file of tags
166 2) Clustering and classification - of given input block groups (from blockbuster tool) as explained in the original paper. 167
168 2) Clustering and classification - of given input blockgroups (output of blockbuster tool) as explained in the original paper.
169
167 3) Post-processing - extracts distribution of clusters searched against Rfam database and plots hierarchical clustering made out of centroids of each BlockClust predicted cluster. 170 3) Post-processing - extracts distribution of clusters searched against Rfam database and plots hierarchical clustering made out of centroids of each BlockClust predicted cluster.
168 171
169 For a thorough analysis of your data, we suggest you to use complete blockclust workflow, which contains all three modes of operation. 172 For a thorough analysis of your data, we suggest you to use complete blockclust workflow, which contains all three modes of operation.
170 173
171 **Inputs** 174 **Inputs**
172 175
173 BlockClust input files are dependent on the mode of operation: 176 BlockClust input files are dependent on the mode of operation:
174 1) Pre-processing mode: 177
175 Binary Sequence Alignment Map (BAM) file 178 1. Pre-processing mode:
176 179 * Binary Sequence Alignment Map (BAM) file
177 2) Clustering and classification: 180
178 A blockgroups file generated by blockbuster tool 181 2. Clustering and classification:
179 Select reference genome 182 * A blockgroups file generated by blockbuster tool
180 183 * Select reference genome
181 3) Post-processing: 184
182 Output of cmsearch, searched clusters generated by BlockClust against Rfam 185 3. Post-processing:
183 BED file containing clusters generated by BlockClust 186 * Output of cmsearch, searched clusters generated by BlockClust against Rfam
184 Pairwise similarities of blockgroups generated by BlockClust 187 * BED file containing clusters generated by BlockClust
185 188 * Pairwise similarities of blockgroups generated by BlockClust
186 **Output** 189
187 1) Pre-processing mode: 190 **Outputs**
188 BED file of tags with expressions 191
189 192 1. Pre-processing mode:
190 2) Clustering and classification: 193 * BED file of tags with expressions
191 Hierarchical clustering plot of all input blockgroups by their similarity 194
192 Pairwise similarities of all input blockgroups 195 2. Clustering and classification:
193 BED file containing predicted clusters 196 * Hierarchical clustering plot of all input blockgroups by their similarity
194 BED file containing prediction of blockgroups by pre-compiled SVM binary classification model. 197 * Pairwise similarities of all input blockgroups
195 198 * BED file containing predicted clusters
196 3) Post-processing: 199 * BED file containing prediction of blockgroups by pre-compiled SVM binary classification model.
197 Distribution of clusters with annotations searched against Rfam database 200
198 hierarchical clustering made out of centroids of each BlockClust predicted cluster 201 3. Post-processing:
202 * Distribution of clusters with annotations searched against Rfam database
203 * Hierarchical clustering made out of centroids of each BlockClust predicted cluster
199 204
200 ------ 205 ------
201 206
202 **References** 207 **References**
203 208