Introduction

Omni-polyA combines several machine learning techniques such as genetic algorithms, deep learning, synthetic minority oversampling, ensemble learning and different classifiers in a tree-like decision structure for deriving a robust classification model for the prediction of poly(A) signals.

The tool receives as an input the DNA sequences of at least 206 nt (shorter sequences will be ignored) in fasta format and will produce a prediction for each conserved hexamer found within the sequence(s). The tool considers the 12 most frequent poly(A) variants (view sample file).

Please cite Magana-Mora A, Kalkatawi M, Bajic VB. Omni-PolyA: a method and tool for accurate recognition of Poly(A) signals in human genomic DNA. BMC Genomics. 2017;18:1:620.