Related Experiment Video
Updated: Jun 4, 2026
08:14
Determining Membrane Protein Topology Using Fluorescence Protease Protection (FPP)
Published on: April 20, 2015
An FPT approach for predicting protein localization from yeast genomic data
Jin Wang1, Chunhe Li, Erkang Wang
1State Key Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun, Jilin, China.
Plos One
|February 2, 2011
Summary
A new frequent pattern tree (FPT) method accurately predicts protein localization, outperforming existing approaches. This data mining technique offers novel insights into protein functions and can guide experimental research.
Area of Science:
- Bioinformatics
- Computational Biology
- Proteomics
Background:
- Accurate protein localization is crucial for understanding cellular functions.
- Genomics and proteomics advancements generate vast datasets requiring sophisticated data mining.
- Protein localization prediction addresses fundamental biological questions.
Purpose of the Study:
- To develop a novel frequent pattern tree (FPT) approach for predicting protein localization.
- To generate a minimum set of rules (mFPT) for accurate protein localization prediction.
- To benchmark mFPT against existing methods and identify novel protein localizations.
Main Methods:
- Developed a frequent pattern tree (FPT) algorithm to derive a minimum set of rules (mFPT).
- Utilized yeast genomic data to acquire prediction rules.
- Compared mFPT performance against Bayesian networks and logistic regression using statistical measures.
Main Results:
- The mFPT approach demonstrated superior prediction accuracy compared to other methods.
- Identified 138 proteins with differing predictions between mFPT and the simple naive Bayesian method (SNB) at a 0.65 hit-rate.
- Generated novel localization predictions for 17 proteins lacking defined annotations.
Conclusions:
- The mFPT method offers a generalized and effective approach for protein localization prediction.
- Novel predictions can guide experimental validation and enrich protein annotation databases.
- The FPT approach has broad applicability to other biological data mining tasks, including interaction and structure-function analyses.
