Bio-medical Ontologies Maintenance and Change Management

More documents

Recommendations

Info

246 C.h. You, L.B. Holder, and D.J. Cook error = #PosEgsNotCvd +#NegEgsCvd #PosEgs +#NegEgs #PosEgsNotCvd is the number of positive examples not containing the substructure, and #NegEgsCvd is the number of negative examples containing the substructure. #PosEgs is the number of positive examples remaining in the experimental set, of which the positive examples that have already been covered in a previous iteration were removed, and #NegEgs is the total number of negative examples, which is constant, because negative examples are not removed. SUBDUE’s supervised learning uses two approaches to minimize error. First, by using the definition of description length SUBDUE tries to find a substructure S minimizing DL(G + |S)+DL(S)+DL(G − ) − DL(G − |S), where the last two terms represent the incorrectly compressed negative example graph. This approach will lead the discovery algorithm toward a larger substructure that characterizes the positive examples, but not the negative examples. In addition to the compression-based evaluation, SUBDUE can use a setcover approach based on the error measure. At each iteration SUBDUE adds a new substructure to the disjunctive hypothesis and removes covered positive examples. This process continues until either all positive examples are covered or no substructure exists discriminating the remain positive examples from the negative examples. 5 Substructure Analysis in Metabolic Pathways Our goal is the application of the SUBDUE graph-based relational learning system to the KEGG metabolic pathways to find better understanding and biologically meaningful substructures. These substructures can distinguish two pathways or provide the common features in several pathways. Research shows that topological features of biological networks are closely related to biological functions [13, 2]. A simple way to apply supervised or unsupervised learning to the pathways is based on molecules, such as genes, proteins and other macro molecules. Because each molecule has a specific structure and other biochemical features, we can easily distinguish two groups or find the common features in a group. But our research is focused on the pattern of the relationship between molecules for system-level understanding of pathways. The pattern of relationship can be shown in a variety of forms, such as biochemical reaction, enzyme activity and signal transduction. This section first introduces our graph representation (section 5.1). As a preliminary task, we describe substructure analysis on individual metabolic pathways (section 5.2). Then we represent our main experiments in this research: supervised learning (section 5.3) and unsupervised learning (section 5.4) on groups of metabolic pathways. The ultimate goal of our exploration
Substructure Analysis of Metabolic Pathways 247 is to show that the substructures found by graph-based relational learning are biologically important and meaningful. 5.1 Graph Representation Input graphs for SUBDUE are converted from KGML files. KGML is a standard data format to express and distribute a biological network from KEGG. There are three major entities in KGML: Entry, Relation and Reaction. Entry represents various biomolecules in the metabolic pathway, such as enzyme, gene, compound and so on. Relation denotes a relationship between two or more enzymes, genes and maps. The maps denote the types of the Entry nodes linked to the other pathways [26]. The names of these Entry nodes represent the name of the linked pathways. Reaction is a biochemical reaction between two or more compounds catalyzed by one or more enzymes. Detailed information on KGML is described in [26]. In biochemical semantics, Entries are nodes of metabolic pathways, and Relations and Reactions are relationships between two or more Entries. enzyme type type S_to_Rct entry reaction name ec:1.4.1.3 name entry E_to_Rct type compound cpd:06560 name reversible E_to_Rel Rct_to_P Rn:R05573 GErel type type relation entry name subtype compound compound cpd:06562 value Rel_to_E S_to_Rct reversible Rn:R05575 type enzyme ec:1.4.1.5 type entry name E_to_Rct type Fig. 3. A graph representation of a metabolic pathway reaction name Rct_to_P entry name compound cpd:06563 In our graph representation, Relations and Reactions are also represented as vertices in order to describe the properties of Relations and Reactions. Vertices representing major entities have two satellite vertices which are connected to their main vertex by edges, labeled as Name and Type, to explain its property. A name vertex linked by the Name edge denotes the KEGG ID, and a type vertex linked by the Type edge describes the property of the entity vertex. A Relation represents the association between two or more Entries (genes or enzymes) by an edge whose label represents a direction from one
Page 1 and 2:
Amandeep S. Sidhu and Tharam S. Dil
Page 3 and 4:
Amandeep S. Sidhu and Tharam S. Dil
Page 5 and 6:
Preface The molecular biology commu
Page 7 and 8:
Preface VII biomedical systems, and
Page 9 and 10:
X Contents Mining Clinical, Immunol
Page 11 and 12:
2 A.S. Sidhu, M. Bellgard, and T.S.
Page 13 and 14:
Page 15 and 16:
Page 17 and 18:
Page 19 and 20:
Towards Bioinformatics Resourceomes
Page 21 and 22:
Page 23 and 24:
Page 25 and 26:
2 The New Waves Towards Bioinformat
Page 27 and 28:
2.4 Semantic Web Towards Bioinforma
Page 29 and 30:
Page 31 and 32:
Page 33 and 34:
Page 35 and 36:
Page 37 and 38:
Page 39 and 40:
Page 41 and 42:
Page 43 and 44:
A Summary of Genomic Databases: Ove
Page 45 and 46:
Page 47 and 48:
Page 49 and 50:
Page 51 and 52:
Page 53 and 54:
Page 55 and 56:
Page 57 and 58:
Page 59 and 60:
Appendix A Summary of Genomic Datab
Page 61 and 62:
Protein Data Integration Problem Am
Page 63 and 64:
Protein Data Integration Problem 57
Page 65 and 66:
Page 67 and 68:
Page 69 and 70:
Fig. 3. Class Hierarchy of Protein
Page 71 and 72:
Page 73 and 74:
Page 75 and 76:
Page 77 and 78:
72 L. Stanescu, D. Dan Burdescu, an
Page 79 and 80:
Page 81 and 82:
Page 83 and 84:
Page 85 and 86:
Page 87 and 88:
Page 89 and 90:
Page 91 and 92:
Page 93 and 94:
Page 95 and 96:
Page 97 and 98:
Page 99 and 100:
Page 101 and 102:
Page 103 and 104:
Page 105 and 106:
100 L. Stanescu, D. Dan Burdescu, a
Page 107 and 108:
Page 109 and 110:
Page 111 and 112:
Page 113 and 114:
Page 115 and 116:
Page 117 and 118:
Page 119 and 120:
Page 121 and 122:
Page 123 and 124:
Page 125 and 126:
Page 127 and 128:
Page 129 and 130:
Page 131 and 132:
Page 133 and 134:
Page 135 and 136:
Page 137 and 138:
Page 139 and 140:
Page 141 and 142:
Page 143 and 144:
Page 145 and 146:
Page 147 and 148:
Bio-medical Ontologies Maintenance
Page 149 and 150:
Page 151 and 152:
Page 153 and 154:
Page 155 and 156:
Page 157 and 158:
Page 159 and 160:
TextToOnto (Maedche and Volz 2001)
Page 161 and 162:
Page 163 and 164:
Page 165 and 166:
Page 167 and 168:
Page 169 and 170:
Page 171 and 172:
Page 173 and 174:
Extraction of Constraints from Biol
Page 175 and 176:
Page 177 and 178:
Page 179 and 180:
Page 181 and 182:
Page 183 and 184:
Page 185 and 186:
Page 187 and 188:
Page 189 and 190:
Page 191 and 192:
Classifying Patterns in Bioinformat
Page 193 and 194:
Page 195 and 196:
Page 197 and 198:
Page 199 and 200: Classifying Patterns in Bioinformat
Page 215 and 216: Mining Clinical, Immunological, and
Page 241 and 242: Substructure Analysis of Metabolic
Page 249: Substructure Analysis of Metabolic
Page 253 and 254: entry compound relation subtype com
Page 257 and 258: Running Time 4500 4000 3500 3000 25
Page 263 and 264: 6 Conclusion Substructure Analysis
Page 267 and 268: 266 J.Y. Chen, S. Taduri, and F. Ll
Page 281 and 282: Completing the Total Wellbeing Puzz
Page 295 and 296: 296 M. Fernandez, M. Villasana, and
Page 301 and 302:
302 M. Fernandez, M. Villasana, and
Page 303 and 304:
Page 305 and 306:
Page 307 and 308:
Page 309 and 310:
Page 311 and 312:
Page 313 and 314:
Page 315 and 316:
Genetic Algorithm in Ab Initio Prot
Page 317 and 318:
Page 319 and 320:
Page 321 and 322:
Page 323 and 324:
Page 325 and 326:
Page 327 and 328:
Page 329 and 330:
Page 331 and 332:
Page 333 and 334:
Page 335 and 336:
Page 337 and 338:
Page 339 and 340:
Page 341:
Author Index Apiletti, Daniele 169
show all

Bio-medical Ontologies Maintenance and Change Management

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?