Abstract | ||
---|---|---|
In the post-genome era, huge numbers of protein structures accumulate, but little is known about their function. It is time consuming and labour intensive to investigate them, e.g., enzyme catalytic properties, through in vivo or in vitro work. So in silico predictions could be a promising strategy to greatly shrink the list of potential targets. This work incorporated both structural and physico-chemical information into a Naive Bayes classification system, and gained much better performance. The ten-fold cross validation results of this method could reach 88.6% of sensitivity and 93.7% of specificity. The improvement of prediction accuracy is detailed in this paper. The PECB is also applied to predict other important sites. |
Year | DOI | Venue |
---|---|---|
2008 | 10.1504/IJBRA.2008.019576 | IJBRA |
Keywords | Field | DocType |
physico-chemical information,post-genome era,naive bayes classification,enzyme catalytic residue,important site,vitro work,better performance,improvement ofprediction accuracy,enzymecatalytic property,insilico prediction,huge number,naive bayesclassification system,enzyme,sequence analysis,bioinformatics,naive bayes | Biology,Naive Bayes classifier,Protein function,Artificial intelligence,Bioinformatics,Cross-validation,Machine learning,In silico,Bayes' theorem | Journal |
Volume | Issue | ISSN |
4 | 3 | 1744-5485 |
Citations | PageRank | References |
1 | 0.36 | 3 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Kunpeng Zhang | 1 | 61 | 12.54 |
Yun Xu | 2 | 167 | 19.13 |
Guoliang Chen | 3 | 305 | 46.48 |