In the annotation file of UCSC, we can see some genes on the + strand and others on the - strand, why? I've always thought DNA's two strand were the same, because the ACGT are paired.
I've always thought DNA's two strand were the same, because the ACGT are paired.
Let's take an example:
5' - . . . A C C G T . . . - 3'(+strand)
3' - . . . T G G C A . . . - 5'(-strand)
Reading from the 5' direction to 3', as this is the direction in which RNA is transcribed, the two strands are distinct (...ACCGT... vs. ...ACGGT...). This results in different transcripts and potentially different genes, depending on the codons.
Just to make sure that there is no misunderstanding on the biology of what is happening because I think that is what the OP is asking. The DNA on the two strands are not the same. They are complementary and as such they will translate to entirely different proteins AAA translates to Lysine whereas its complement TTT will be translated to Phenylalanine
The second essential ingredient is the directionality of the DNA as it can be transcribed in only one direction.
The two strands are actually not the same. One is the reverse complement of the other. If a polymerase would read AAATCC on one strand of a 6bp DNA fragment, it would read GGATTT on the other strand.
Probably the simplest case (I'm sure there are caveats):
The exons in gene+ do not overlap with the exons in gene-.
+ATG GGG TGA-
-AGT TGG GTG GTA+
As zev points out so far we know that genes in one strand do not overlap with genes in another strand, having genes in both strands increases the number of genes an organism can have without having a disproportionally larger genome.
That said, there might be the case that two genes might perfectly overlap, there's not a physical constraint as long as both sequences carry a meaning in translation. It just hasn't been observed so far.
Thank you for all your answers. I've thought that the genes came from the meaning strand, and the other strand didn't code any genes. It seems that the understanding of the two strands were not correct.
The process of transcription is somehow random, and each of the strands can be recognized as the template of genes. Is it right?
Human chromosome 15 is about 100 millions bases long. Some genes are transcribed on one side , some on the other. Genes almost never overlap (some viruses might be exceptions)
Transcription is not random. Usually, there are recognition sequences upstream of gene starts, and if the right protein is sitting on them, the transcription mechanism goes to work close by. But in many cases, if you flipped a gene (and it's associated controlling sequences) around, everything would still be fine.
Binding of transcription factors and histone modifications that open up chromatin are what generally drive the transcription of various genes. It's not so much random, but stochastic, where a population of factors will "bounce around" until they find energetically-favorable binding sites, or chemical modifications are made by enzymes to a certain population of histones, etc. which activate or shut down different parts of the genetic program.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
I think they can and do overlap sometimes. e.g. "Overlapping genes are known to be common in viruses, mitochondria, bacteria, and plasmids [1], but are thought to be rare in eukaryotes." from here: http://www.biomedcentral.com/1471-2164/9/169 I'm sure there are better references, this was just my first Google hit.