INSWF DNA signal analysis tool: intelligent noise suppression window filter

Muneer Ahmad, Iftikhar Ahmad, Muhammad Bilal, Alireza Jolfaei, Raja Majid Mehmood*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


DNA signals mainly differ from standard digital signals due to their biological data contents. Owing to unique properties of DNA signals the conventional signal processing techniques, such as digital filters, suffers with spectral leakage and results in insignificant noise suppression in DNA sequence analysis. This article presents an intelligent noise suppression window filter (INSWF) for DNA signal analysis. The filter demises the signal by separating high-level frequency contents and by identifying nucleotides with high fuzzy membership contribution at particular locations. The nucleotide contents of signals are later filtered by application of median filtering employing a combination of s-shaped and z-shaped filters. The fundamental characteristic of codons usage that causes uneven nucleotides segmentation has been tackled by finding the best fit of the curve in biological contents of filter. One of the fuzzy correlations existing between codons and median that nucleotides incorporated to reduce the signal noise to a larger magnitude. The INSWF filter outperformed the existing fixed-length digital filters tested over 250 benchmarked and random datasets of various species. A notable enhancement of 45% to 130% was achieved by significantly suppressing signal noise as compared with conventional digital filters in DNA sequence analysis.

Original languageEnglish
Pages (from-to)670-685
Number of pages16
JournalSoftware - Practice and Experience
Issue number3
Early online date24 Aug 2020
Publication statusPublished - Mar 2021


  • adaptive digital filter
  • codon usage
  • digital filter
  • DNA sequence analysis
  • fixed-length filter
  • fuzzy rules
  • signal noise


Dive into the research topics of 'INSWF DNA signal analysis tool: intelligent noise suppression window filter'. Together they form a unique fingerprint.

Cite this