The road between gene and final protein product is long and not without unexpected turns. In extreme cases, the computer translation of the gene sequence bears as little resemblance to the mature protein as tadpoles to frogs. In this issue of Cell, the groups of Yoshida and Kan report a new turn in this road, synthesis of a fused protein. Before we summarize the experiments on which the authors base their claim, we briefly review here the other processes that are known to modify the original genetic message. Although RNA synthesis can be counted on to yield a faithful copy of the template, occasional programmed slips do occur. Thus far, these have been limited to expansion of purine homopolymer runs (see Thomas et al., 1966); they probably occur by back-slippage (“stuttering”) of the polymerase on the template. Only in paramyxoviruses does such an insertion of nucleotides change the amino acid sequence of the protein product (Thomas et al., 1966; Cattaneo et al., 1969). Four types of RNA processing may change the original message, and at present, none can be faithfully predicted from the genomic sequence. l RNA splicing. Recognition of RNA splice sites is not only dependent on RNA primary structure and folding, but also on proteins that interact with the RNA, as illustrated by the profusion of cell-specific alternative splicing pathways (Breitbart et al., 1967). l Tmns-splicing is a variation on the &-splicing theme (see Van der Ploeg, 1966). Most examples of franssplicing, e.g., in trypanosomes and nematodes, do not affect the proteincoding sequences of the pre-mRNA. This is the case, however, for the trans-splicing observed for some chloroplast RNAs (Choquet et al., 1966). l RNA editing of pre-mRNAs leads to insertion or removal of U’s (Benne, 1969; Simpson and Shaw, 1969). Recent work indicates that editing is posttranscriptional, and it seems increasingly unlikely that an RNA minus strand is involved. Presumably the editing machinery infers from the pre-mRNAs where the (extensive) alterations are required, but scrutiny of the available sequences has not yet given a clue as to how this is done. To date, RNA editing has only been observed in the mitochondria of trypanosomes and other kinetoplastids. l C+J conversion. The only known example of this process occurs in the pm-mRNA for apolipoprotein B, in Minireview