vaccineda

vaccineda predicts immunomodulatory oligodeoxynucleotide (IMODN) sequences and supports the design of DNA-based vaccine adjuvants by applying support vector machine models trained on experimentally validated IMODNs to identify sequence patterns, including unmethylated CpG motifs that activate toll-like receptor 9 (TLR9).

Key Features:

• Experimentally validated dataset: Models were developed from analysis of 2,193 experimentally validated IMODNs curated from the literature.
• Nucleotide composition analysis: The tool analyzes nucleotide and oligonucleotide composition, including pentanucleotide features, to characterize IMODNs.
• Frequent motif identification: Analysis revealed prevalent motifs/compositions such as T, GT, TC, TT, CGT, TCG, and TTT in effective IMODNs.
• Support Vector Machine (SVM) models: SVM-based predictive models were trained on composition features to classify IMODNs.
• Pentanucleotide composition performance: Pentanucleotide composition models achieved an MCC of 0.75 and accuracy of 87.57%.
• Motif-integrated models: Integration of motif information increased performance to an MCC of 0.77.
• Palindromic IMODN models: Specialized models for palindromic IMODNs achieved an MCC of 0.84 and accuracy of 91.94%.
• Model validation: Models were evaluated using five-fold cross-validation and validated on an independent dataset.

Scientific Applications:

• Design of IMODN-based vaccine adjuvants: Prediction and selection of immunomodulatory oligodeoxynucleotides for inclusion in DNA-based adjuvant formulations.
• Sequence optimization for TLR9 activation: Identification of nucleotide compositions and motifs, including unmethylated CpG patterns, to enhance TLR9-mediated innate immune activation.
• Prediction of palindromic IMODNs: Detection and prioritization of palindromic sequences with higher predictive performance for immunomodulatory activity.

Methodology:

Analysis of 2,193 experimentally validated IMODNs; calculation of nucleotide and oligonucleotide (including pentanucleotide) composition features; motif analysis identifying frequent motifs (T, GT, TC, TT, CGT, TCG, TTT); development of support vector machine models; integration of motif information; separate SVM models for palindromic IMODNs; evaluation by five-fold cross-validation and validation on an independent dataset.