"some researchers now think that *de novo* genes could be quite common."

The definition of a gene is any unit factor that is inherited from the parents.  Today that definition also includes the limiter that the factor is carried on the DNA.  We went thru a period in the 1960s and 70s where another limiter was also added which said genes only produce proteins or towards the end of that period RNA as their end products.  But we have since found all kinds of unit factors inherited from the parents that did not produce either RNA or proteins.  In fact many of those unit factors produced nothing at all.  They simply changed the way some protein producing gene behaved.  It could be as simple as causing the protein producing gene to produce more or less protein.  Or the protein producing gene might change the place in the cell where the protein did its job.  Or it might express protein in tissue that it did not express in before or fail to express in tissues it should express in.  Genes that do not produce either a protein or RNA are often called non coding genes.  Even a deletion in the non coding DNA can result in a gene that expresses as a unit character.  There are probably two or three non coding genes for every coding gene.  Possibly even more.  How many there are depends a lot on how you define alleles.  Making non coding genes "de novo" is dirt simple.

The authors of the paper Peter linked to are talking about "de novo" coding genes.  Making such genes is much harder than making non coding genes.  You not only must have a coding sequence that produces a useful protein but you must also have both start and stop sequences plus possibly a promoter or switch of some type to allow that coding section to express.  It is nearly inevitable that some of this extra required DNA is going to be stolen or borrowed from other coding genes or from some pseudo gene that codes for a useless protein that has no biological function.  There are lots of pseudo genes.  Many result from horizontal gene transfer from other species.  The last count in humans that I saw was something like 15,000 pseudo genes, so nearly as many as actual useful coding genes.

Dick

             ***********************************************
The BEE-L mailing list is powered by L-Soft's renowned
LISTSERV(R) list management software.  For more information, go to:
http://www.lsoft.com/LISTSERV-powered.html