T cells scan the surface of antigen presenting cells (APCs) to detect foreign peptides on major histocompatibility complex (MHC). Recognition takes place via the binding of T cell receptor (TCR) to peptide-MHC (pMHC) and the transduction of downstream signaling cascades. To induce a proper immune response, it is essential that T cells are activated when they bind to foreign peptides and not when confronted with self-peptides. Although T cells that have high affinity to self-peptides are deleted during development in the thymus, T cells in the periphery should still discriminate foreign peptides with high affinity to its TCR from the many and abundant low affinity self peptides. Failures in this discrimination result in either infection or autoimmune diseases.
     Many models have been suggested to explain this, however, it is not clear that how peptide discrimination by TCRs is disturbed in the background of many low affinity self-peptides that could antagonize signaling. Here we simulate the TCR signaling model developed by Francois et al. 2013 with low affinity ligands, and quantify the ability of ligand discrimination using mutual information. The mutual information indicates the maximum number of input signal values that a signaling pathway can reliably resolve. The results show that ligand discrimination works only in the presence of antagonists with very low affinity, which suggests that negative selection of immature T cells should be very strict. We will also discuss how the selection of immature T cells with different strength of TCR signaling results in different T cell differentiation.