[Paper Review] Computational prediction of replication sites in DNA sequences using complex number representation

Shubham Kundal, Raunak Lohiya+4 more|arXiv (Cornell University)|Sep 27, 2019

RNA and protein synthesis mechanisms27 references5 citations

TL;DR

This paper proposes iCorr, a novel computational method that uses complex number representation of DNA sequences to predict origin of replication (ORI) sites with higher resolution and accuracy than existing auto-correlation methods. By mapping nucleotides to complex units {+1, −1, +i, −i}, iCorr captures full spatial information of all four bases, enabling automatic peak detection for ORI localization without manual inspection, significantly improving prediction precision in both prokaryotes and eukaryotes.

ABSTRACT

Computational prediction of origin of replication (ORI) has been of great interest in bioinformatics and several methods including GC-skew, auto-correlation etc. have been explored in the past. In this paper, we have extended the auto-correlation method to predict ORI location with much higher resolution for prokaryotes and eukaryotes, which can be very helpful in experimental validation of the computational predictions. The proposed complex correlation method (iCorr) converts the genome sequence into a sequence of complex numbers by mapping the nucleotides to {+1,-1,+i,-i} instead of {+1,-1} used in the auto-correlation method (here, i is square root of -1). Thus, the iCorr method exploits the complete spatial information about the positions of all the four nucleotides unlike the earlier auto-correlation method which uses the positional information of only one nucleotide. Also, the earlier auto-correlation method required visual inspection of the obtained graphs to identify the location of origin of replication. The proposed iCorr method does away with this need and is able to identify the origin location simply by picking the peak in the iCorr graph.

Motivation & Objective

• To overcome limitations of existing auto-correlation methods in predicting origin of replication (ORI) in DNA sequences.
• To eliminate the need for visual inspection of correlation profiles in ORI prediction.
• To enhance resolution and accuracy by leveraging full spatial information of all four nucleotides (A, T, C, G) using complex number mapping.
• To enable more precise experimental validation by reducing search space through higher-resolution predictions.
• To provide a robust, automated method applicable to both prokaryotic and eukaryotic genomes.

Proposed method

• The iCorr method maps each nucleotide to a complex number: A → +1, T → −1, C → +i, G → −i.
• This transforms the DNA sequence into a complex-valued sequence, preserving positional and compositional information of all four bases.
• The method computes a complex auto-correlation function (iCorr) over sliding windows of the genome sequence.
• The iCorr function is calculated using the standard auto-correlation formula applied to the complex sequence, capturing phase and amplitude information.
• The ORI location is identified automatically as the peak in the iCorr profile, eliminating manual interpretation.
• The method uses a sliding window approach with adjustable window and shift sizes to optimize resolution and noise reduction.

Experimental results

Research questions

• RQ1 Can complex number representation of DNA sequences improve the resolution and accuracy of origin of replication prediction compared to traditional auto-correlation methods?
• RQ2 Can the iCorr method automatically detect ORI locations without requiring visual inspection of correlation profiles?
• RQ3 Does incorporating positional information of all four nucleotides (A, T, C, G) via complex mapping enhance detection sensitivity and specificity?
• RQ4 How does the iCorr method perform across diverse prokaryotic and eukaryotic genomes in terms of ORI prediction accuracy?
• RQ5 What optimal window and shift sizes maximize prediction precision in the iCorr framework?

Key findings

• The iCorr method successfully identifies the origin of replication as the peak in the complex correlation profile, eliminating the need for manual inspection required in traditional auto-correlation methods.
• By using complex number mapping, iCorr captures full spatial information of all four nucleotides, unlike previous methods that only use G nucleotide positions.
• The method achieves higher resolution prediction of ORI locations, which is critical for guiding experimental validation.
• For *Bacillus subtilis*, iCorr, gCorr, and GC-skew all predict the terminus (TER) region near the same genome position, but iCorr provides a sharper, more distinct peak.
• The iCorr method is applicable to both prokaryotes and eukaryotes, with consistent performance in identifying ORI locations without requiring genome-specific tuning.
• Optimal window and shift sizes were identified through analysis, demonstrating that parameter optimization improves prediction precision.