SwarmTCR

a computational approach to predict the specificity of T Cell Receptors. SwarmTCR predicts T-cell receptor (TCR) specificity using the 'distance' between TCRs. SwarmTCR is an optimized adaption of TCRdist (Dash et al. 2017) that uses both protein sequence identity of the complementary determining region (CDR) loops and particle swarm optimization. Distance is determined in the same manner as the original TCRdist methodology, where the alignment (BLOSUM62 matrix) values at each CDR loop are multiplied by a weight, the sum of all CDR loop values is the distance. Unlike TCRdist, SwarmTCR optimizes the weight for each CDR loop, making it specific to the repertoire tested. SwarmTCR is functional on both single-cell and deep-sequencing data, input parameters for each will be detailed below.