DivergentSet, a tool for picking non-redundant sequences from large sequence collections

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Selecting representative sequence subsets is crucial for bioinformatics analyses.
• Existing methods for constructing divergent sets can be computationally intensive.
• The stability of identified biological motifs is often dependent on sequence selection.

Purpose of the Study:

• To introduce DivergentSet, a novel bioinformatics tool for constructing representative sequence sets.
• To demonstrate the computational efficiency of using phylogenetic trees for divergent set construction.
• To highlight the utility of DivergentSet in enhancing the reliability of motif discovery.

Main Methods:

• Development of the DivergentSet software with a user-friendly interface.
• Utilizing phylogenetic trees to guide the selection of divergent sequences.
• Comparison of DivergentSet's performance against a naive distance matrix method.
• Application of DivergentSet to analyze motif stability in biological sequences.

Main Results:

• DivergentSet construction using phylogenetic trees is up to 100 times faster than naive methods.
• The software integrates sequence retrieval, set refinement, and random set generation.
• Motif analysis using MEME (Motif Elicitation by Maximum Entropy) shows high instability based on sequence choice.
• DivergentSet facilitates sensitivity analyses for motif discovery.

Conclusions:

• DivergentSet offers a computationally efficient and user-friendly solution for selecting representative sequence sets.
• The tool enhances the robustness of bioinformatics analyses, particularly in motif and covariation identification.
• Sensitivity analyses enabled by DivergentSet are vital for identifying biologically significant motifs.