Gene birth contributes to structural disorder encoded by overlapping genes

S. Willis; J. Masel

Some homologous pairs had such dissimilar protein sequences that ISD values were essentially independent. We therefore manually analyzed sequence similarity within each homology group using the Geneious [18] aligner with free end gaps, using Blosum62 as the cost matrix. The percent similarity using the Blosum62 matrix with similarity threshold 1 was then used as the criterion for whether a protein sequence would remain in its homology group for the ISD analysis. We used ≥ 50% protein sequence similarity as the threshold to assign a link between a pair, and then used single-link clustering to assign protein sequences to 561 distinct homology groups. Pre-overlapping genes were then assigned to the homology groups of the corresponding ancestral genes.