werner@aecom.YU.EDU (Robert N. Berlinger) (06/14/88)
As a means of teaching someone how to use various DNA sequence analysis programs, I picked a gene at random out of Genbank (Actually, it was really pseudorandom -- I picked a gene that I had a good story about, the human Alcohol Dehydrogenase gene, and I picked an entry that was rather short, entry HUMADHIB, about 130 base pairs. Then I made a reverse complement, just to trick the person up, and ran it through an otherwise useless PostScript program of my own making which, given a sequence, recreates the Sequencing gel autoradiogram. He read the gel in, made two mistakes, a frameshift and a missence, and did not check for open reading frames, so he didn't undo the reverse complemement that I had nefariously performed. He ran the comparison against all of Genbank. The result was astounding: the reverse complement of Human ADH Class I Beta, WITH those two mistakes is in 130 base pairs 100% similar to part of the intergenic region of pBR322. It absolutely blew me away. If someone had told me that two pieces of essentially unrelated DNA could be identical over 130 bases I would have said that they were crazy, but I saw it with my own eyes. Amazing. -- Craig Werner (future MD/PhD, 4 years down, 3 to go) werner@aecom.YU.EDU -- Albert Einstein College of Medicine (1935-14E Eastchester Rd., Bronx NY 10461, 212-931-2517) "The end. 94. 95. The very, very, very end."
eddy@boulder.Colorado.EDU (Sean Eddy) (06/14/88)
In article <1843@aecom.YU.EDU> werner@aecom.YU.EDU (Robert N. Berlinger) writes: > The result was astounding: the reverse complement of Human >ADH Class I Beta, WITH those two mistakes is in 130 base pairs 100% >similar to part of the intergenic region of pBR322. It absolutely >blew me away. If someone had told me that two pieces of essentially >unrelated DNA could be identical over 130 bases I would have said that >they were crazy, but I saw it with my own eyes. Amazing. In fact, I'm so amazed I'd be tempted to find out how ADH class I beta was sequenced. Like, for instance, was it in (or subcloned from) a pBR vector?? 130 bp of 100% sequence similarity suggests a major muck-up... - Sean Eddy - Molecular/Cellular/Developmental Biology; U. of Colorado at Boulder - eddy@boulder.colorado.EDU !{hao,nbires}!boulder!eddy - - "That such pygmies should cast such giant shadows only shows how - late in the day it has become." - - biochemist Erwin Chargaff, of Watson & Crick
wrp@biochsn.acc.virginia.edu (William R. Pearson) (06/14/88)
]about, the human Alcohol Dehydrogenase gene, and I picked an entry ]that was rather short, entry HUMADHIB, about 130 base pairs. Then ] He ran the comparison against all of Genbank. ] ] The result was astounding: the reverse complement of Human ]ADH Class I Beta, WITH those two mistakes is in 130 base pairs 100% ]similar to part of the intergenic region of pBR322. It absolutely ]blew me away. If someone had told me that two pieces of essentially ]unrelated DNA could be identical over 130 bases I would have said that ]they were crazy, but I saw it with my own eyes. Amazing. ] What you saw with your own eyes was probably a mistake in Genbank, not a miraculous sequence similarity. I checked the latest release of Genbank (release 55.0), and could not find a HUMADHIB. There were, however, HUMADH1B[1-7], seven exons for the gene. None of these seven exons has any significant similarity to pbr322, in either orientation. I suspect a piece of pbr322 crept in to Genbank in the wrong place, and has since been removed. Bill Pearson wrp@virginia.EDU