POLYAR: A New Tool for Predicting Poly(A) Sites in Human Sequences
Author Information
Author(s): Akhtar Malik Nadeem, Bukhari Syed Abbas, Fazal Zeeshan, Qamar Raheel, Shahmuradov Ilham A
Primary Institution: COMSATS Institute of Information Technology, Islamabad, Pakistan
Hypothesis
Can a new computer program improve the prediction of poly(A) sites in human mRNA?
Conclusion
POLYAR shows high accuracy in predicting PAS-strong poly(A) sites, but its effectiveness for PAS-weak and PAS-less sites is limited.
Supporting Evidence
- POLYAR achieved a sensitivity of 80.8% for PAS-strong sites.
- POLYAR outperformed the previous best tool, polya_svm, in specificity for PAS-strong sites.
- The study identified 28761 human mapped poly(A) sites classified into three categories.
Takeaway
The POLYAR program helps find important parts of genes that tell the cell where to add a special tail to RNA, which is needed for making proteins.
Methodology
The study involved training and testing datasets of human mapped poly(A) sites using linear discriminant analysis to classify true and false poly(A) sites.
Limitations
The program's accuracy for PAS-weak and PAS-less sites is low, indicating gaps in understanding the factors involved in polyadenylation.
Statistical Information
P-Value
p<0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website