My genome has how many genes?
Here’s an answer: 22,333 – maybe. But no one really knows for sure.
That’s the answer that bioinformatician Steven Salzberg from the University of Maryland gave this evening (Monday, 11 October) in a keynote lecture at Beyond the Genome, a conference running in Boston, Massachusetts.
The figure Salzberg said he favors is one proposed by a public database of DNA sequences called RefSeq. But not all of those genes have been individually confirmed, he noted, and other databases have proposed other figures generally falling in about the 20,000 range, give or take a few thousand.
It’s not a simple question. Researchers generally define a gene as a protein-coding piece of DNA, with regulatory elements on each end, but “we don’t know how many proteins are produced by each gene,” said Salzberg. Increasingly, studies are showing that most single genes can be transcribed to produce different snippets of RNA. “So the number of proteins may be much greater” than the gene count.
This ambiguity may be why the proposed gene count has changed so dramatically since researchers first began thinking about the question more than 20 years ago. At the time, the number was estimated at around 100,000 - a figure thought to have come from a back-of-the-envelope calculation made by Walter Gilbert, a physicist-turned-molecular biologist and co-recipient of the 1980 Nobel Prize in Chemistry, along with Frederick Sanger and Paul Berg, for devising techniques to sequence DNA. Gilbert supposedly proposed it in response to a reporter’s question in 1985 or so, and it stuck until the human genome was finally sequenced in 2000, Salzberg said. At that point, the estimate plunged to about 40,000, and has been steadily pared down since.
Of course, the elephant in the room is, who cares exactly how many genes there are? In the Q&A period after the talk, Julie Segre of the National Human Genome Research Institute raised just this question. “What if there is no set number?” she asked – what if the more relevant thing to count is the RNA and proteins produced?
Gilbert himself liked to say that as a physicist by training, he was glad to have gotten the number in the right order of magnitude, Segre said, suggesting that precision may not be achievable on this particular matter.
“That may be, but I’m a computer scientist by training,” Salzberg replied, to laughter from the audience. “And I don’t like those kinds of approximations.”
Undoubtedly, some interesting ideas will emerge tomorrow at a session dedicated to the true gene count.
That’s the answer that bioinformatician Steven Salzberg from the University of Maryland gave this evening (Monday, 11 October) in a keynote lecture at Beyond the Genome, a conference running in Boston, Massachusetts.
The figure Salzberg said he favors is one proposed by a public database of DNA sequences called RefSeq. But not all of those genes have been individually confirmed, he noted, and other databases have proposed other figures generally falling in about the 20,000 range, give or take a few thousand.
It’s not a simple question. Researchers generally define a gene as a protein-coding piece of DNA, with regulatory elements on each end, but “we don’t know how many proteins are produced by each gene,” said Salzberg. Increasingly, studies are showing that most single genes can be transcribed to produce different snippets of RNA. “So the number of proteins may be much greater” than the gene count.
This ambiguity may be why the proposed gene count has changed so dramatically since researchers first began thinking about the question more than 20 years ago. At the time, the number was estimated at around 100,000 - a figure thought to have come from a back-of-the-envelope calculation made by Walter Gilbert, a physicist-turned-molecular biologist and co-recipient of the 1980 Nobel Prize in Chemistry, along with Frederick Sanger and Paul Berg, for devising techniques to sequence DNA. Gilbert supposedly proposed it in response to a reporter’s question in 1985 or so, and it stuck until the human genome was finally sequenced in 2000, Salzberg said. At that point, the estimate plunged to about 40,000, and has been steadily pared down since.
Of course, the elephant in the room is, who cares exactly how many genes there are? In the Q&A period after the talk, Julie Segre of the National Human Genome Research Institute raised just this question. “What if there is no set number?” she asked – what if the more relevant thing to count is the RNA and proteins produced?
Gilbert himself liked to say that as a physicist by training, he was glad to have gotten the number in the right order of magnitude, Segre said, suggesting that precision may not be achievable on this particular matter.
“That may be, but I’m a computer scientist by training,” Salzberg replied, to laughter from the audience. “And I don’t like those kinds of approximations.”
Undoubtedly, some interesting ideas will emerge tomorrow at a session dedicated to the true gene count.
0 comments:
Post a Comment