Clustering protein sequences with a novel metric transformed from sequence similarity scores and sequence alignments with neural