Cross-species single-cell transcriptomic analysis using evolutionarily informed gene sets

This repository contains the bi-directional BLASTp protein sequence alignment results used to construct the cross-species gene homology graph in the study [Evolutionarily informed gene sets reveal conserved and lineage-modified transcriptional programs in vertebrate forebrain evolution]. Pairwise BLASTp alignments were performed between all species pairs using protein sequences derived from NCBI RefSeq proteomes. For each pair of proteins from different species, alignments were computed in both directions. When multiple high-scoring segment pairs were identified, the alignment with the highest bit score was retained. Only alignments with an E-value < 1e-6 were included. These results were used to define reciprocal homology relationships and construct a weighted multipartite graph, where nodes represent genes and edges represent sequence similarity between homologous proteins. Edge weights correspond to the maximum reciprocal bit score for each gene pair. This dataset enables: (1) reconstruction of cross-species gene homology graphs; (2) identification of homologous gene sets using graph-based community detection; (3) comparative analysis of gene families across vertebrate evolution. The dataset includes BLASTp outputs for all pairwise species comparisons across eleven vertebrate species spanning ~500 million years of evolution, from sea lamprey to human.