In silico data mining of human body fluids to unravel the immunomes in breast cancer

Breast cancer is asymptomatic with a poor prognosis at late stage. We used in-silico based proteomics data mining to identify immunoproteins in body fluids as potential indicators of onset and progression of breast cancer. We curated a list of differentially expressed proteins in tissue and body fluids (e.g. saliva and blood) on subtype and non-subtype specific breast cancer over the last 20 years, to show the extent of similarities in protein expression in tissue and bio-fluids. Based on specific selection criteria, significantly altered proteins were filtered to functionally annotate and analyze using different databases. Furthermore, the immunoproteins that showed cross-talks were further analyzed for amino-acid sequence-specific alterations associated with breast cancer to predict their potential impact on the disease. The curated datasets consolidated 4716 non-redundant proteins collectively in tissue, blood, and saliva from literature focused on subtype or non-subtype specific breast cancer. Of these immunoproteins, 39 (e.g., saliva) and 20 (e.g., blood) were found to cross-talk, of which 28 and 6 from saliva and blood, respectively, showed amino acid variations and were associated with breast cancer. Similarly, a total of 92 and 10 driver mutants were identified in saliva and blood, respectively, with ‘deleterious’ or ‘damaging’ impact on the biological function of a protein. The results of the study established correlations between expression profile and variation of immunoproteins with breast cancer to assess the cumulative effect of mutational hotspots and identify proteome-scale alterations that could trigger abnormal cell behavior.