Fig. 1: Antibody epitope analysis using programmable phage display of peptide libraries.
From: Efficient encoding of large antigenic spaces by epitope prioritization with Dolphyn
A Complexity and reactivity of previously published peptide libraries. Bars show number of peptides included in each library and the percentage of peptides that are reactive in at least 1% of samples from the Vaccine Research Center (VRC) cohort. B The Public Epitope Data Set includes a k-mer scan and an alanine scan of 544 virus-derived 56 amino acid long immunodominant peptides in 59 individuals (425 samples). The k-mer scan consists of k = 15–45 amino-acid-long sub-peptides of the wildtypes, starting every 5 amino acids. The alanine scan consists of modified versions of the wildtype peptide where triplets of amino acids were replaced with three alanines. Wildtype alanines are replaced by glycines. C Compilation of alanine scans from reactive peptides and individuals. Each gray line is the difference of the alanine peptide at that position to the wildtype reactivity in one individual. Only lines indicating a single epitope were included and shifted to the center. Blue line indicates median. D Summarized k-mer scans. A peptide is considered reactive if more than one percent of samples react to it. The Training Set indicates those peptides used in the prediction model introduced in Fig. 2. | Source data are provided as a Source Data file.
