TY - JOUR
T1 - Prediction of EST functional relationships via literature mining with user-specified parameters
AU - Wang, Hei Chia
AU - Huang, Tian Hsiang
N1 - Funding Information:
Manuscript received May 9, 2008; revised October 27, 2008. First published December 2, 2008; current version published May 6, 2009. This work was supported by the National Science Council, Taiwan, under Grant NSC 95-2221-E-006-161 and Grant NSC 95-2627-B-006-002. Asterisk indicates corresponding author. *H.-C. Wang is with the Institute of Information Management, National Cheng Kung University, Tainan 701, Taiwan (e-mail: [email protected]. edu.tw).
PY - 2009/4
Y1 - 2009/4
N2 - The massive amount of expressed sequence tags (ESTs) gathered over recent years has triggered great interest in efficient applications for genomic research. In particular, EST functional relationships can be used to determine a possible gene network for biological processes of interest. In recent years, many researchers have tried to determine EST functional relationships by analyzing the biological literature. However, it has been challenging to find efficient prediction methods. Moreover, an annotated EST is usually associated with many functions, so successful methods must be able to distinguish between relevant and irrelevant functions based on user specifications. This paper proposes a method to discover functional relationships between ESTs of interest by analyzing literature from the Medical Literature Analysis and Retrieval System Online, with user-specified parameters for selecting keywords. This method performs better than the multiple kernel documents method in setting up a specific threshold for gathering materials. The method is also able to uncover known functional relationships, as shown by a comparison with the Kyoto Encyclopedia of Genes and Genomes database. The reliable EST relationships predicted by the proposed method can help to construct gene networks for specific biological functions of interest.
AB - The massive amount of expressed sequence tags (ESTs) gathered over recent years has triggered great interest in efficient applications for genomic research. In particular, EST functional relationships can be used to determine a possible gene network for biological processes of interest. In recent years, many researchers have tried to determine EST functional relationships by analyzing the biological literature. However, it has been challenging to find efficient prediction methods. Moreover, an annotated EST is usually associated with many functions, so successful methods must be able to distinguish between relevant and irrelevant functions based on user specifications. This paper proposes a method to discover functional relationships between ESTs of interest by analyzing literature from the Medical Literature Analysis and Retrieval System Online, with user-specified parameters for selecting keywords. This method performs better than the multiple kernel documents method in setting up a specific threshold for gathering materials. The method is also able to uncover known functional relationships, as shown by a comparison with the Kyoto Encyclopedia of Genes and Genomes database. The reliable EST relationships predicted by the proposed method can help to construct gene networks for specific biological functions of interest.
UR - http://www.scopus.com/inward/record.url?scp=67349119041&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=67349119041&partnerID=8YFLogxK
U2 - 10.1109/TBME.2008.2009765
DO - 10.1109/TBME.2008.2009765
M3 - Article
C2 - 19272867
AN - SCOPUS:67349119041
SN - 0018-9294
VL - 56
SP - 969
EP - 977
JO - IEEE Transactions on Biomedical Engineering
JF - IEEE Transactions on Biomedical Engineering
IS - 4
M1 - 4694115
ER -