In silico design and performance of peptide microarrays for breast cancer tumour auto-antibody testing

摘要: The simplicity and potential of minimally invasive testing using sera from patients makes auto-antibody based biomarkers a very promising tool for use in cancer diagnostics. Protein microarrays have been used the identification such signatures. Because high throughput protein expression purification is laborious, synthetic peptides might be good alternative microarray generation multiplexed analyses. In this study, we designed 1185 antigenic peptides, deduced proteins expressed by 642 cDNA clones found to sero-reactive both breast tumour controls. corresponding were production peptide microarrays. Serum samples females with benign malignant tumours healthy control ( n =16 per group) then analysed. Correct classification serum on 78% discrimination ‘malignant versus controls’, 72% ‘benign malignant’ 94% controls’. On arrays, correct these contrasts was 69%, 59% 59%, respectively. over-representation analysis classifiers derived class prediction showed enrichment genes associated ribosomes, spliceosomes, endocytosis pentose phosphate pathway. Sequence analyses highest sero-reactivity demonstrated zinc-finger domain. Peptides’ sero-reactivities negatively correlated hydrophobicity positively positive charge, inter-residue contact energies secondary structure propensity bias. This study hints at possibility silico as an improvement