| Autor: |
Dea Siska, Saskya Mary Soemartojo, Titin Siswantining, Devvi Sarwinda |
| Rok vydání: |
2020 |
| Předmět: |
|
| Zdroj: |
2020 4th International Conference on Informatics and Computational Sciences (ICICoS). |
| DOI: |
10.1109/icicos51170.2020.9299101 |
| Popis: |
Triclustering is the expansion of clustering and biclustering methods that works on three-dimensional (3D) data. This method is generally implemented in the analysis of 3D gene expression data to find gene expression profiles. This data consists of three dimensions: genes, experimental conditions, and time points. Triclustering can group these dimensions simultaneously and form a 3D cluster called a tricluster. Order Preserving Triclustering (OPTricluster) is a triclustering algorithm that uses a pattern-based approach and is used to analyze short time-series data (3–8 time points). The OPTricluster forms the tricluster by identifying genes with the same pattern of change in expression across time points under several experimental conditions. In contrast to most triclustering algorithms that only focus on similarities between experimental conditions, OPTricluster considers the similarities and differences between them. In this study, OPTricluster was implemented with several scenarios in gene expression data of yellow fever patients after vaccination. The lowest average Tricluster Diffusion (TD) score indicates the scenario with the best triclustering result. For this case, we found that the scenario with threshold of 1.6 is the scenario that produced triclusters with better quality (lowest average TD score) than the other scenarios. These triclusters represent gene expression profiles that show the biological relationship among those patients, including anomalies found in patients. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
