Método in silico para análise de sequências de imunoglobulinas produzidas por tecnologia de phage display

Since high-throughput sequencing (HTS) platforms provide larger sampling of phage display libraries, the amount of data imposes challenges to analyze libraries diversity and to find selected clones, which are traditionally tested by antibody affinity assays. Considering that, we developed an automated in silico method to analyze immunoglobulin sequences produced by phage display, which allows the detection of selected clones, from libraries sequenced by HTS platforms. The method consists of 6 steps: reads joining, sequence filtering, translation, enrichment analysis, residues numbering and germline classification. In order to validate the method, 3 sets of data were analysed, each containing initial and final phage display libraries, being 2 sets sequenced by Illumina and one by 454 Roche platform. The complete analysis of each pair of libraries was performed in less than 3 hours. The promising execution time is mainly due to the translation and frequency calculation programs, which were developed with intelligent strategies to process libraries composed of more than 106 reads, in less than 5 minutes. As final output, the method creates a list of candidate clones, enriched and recognized as immunoglobulin variable domain, sorted by fold change of frequency and classified by germline, which probably were selected by phage display experiments. Besides the efficiency of the method concerning the fast performance, the present approach uses a biological criterion to find candidate clones, based on canonical signature of immunoglobulin variable domain.