There’s an old joke about a drunk who lost his keys coming home from the bar one night. Daybreak found him still crawling around searching under a lamppost. A passerby asked him why he didn’t look elsewhere. Drunk’s answer: “That’s where the light is”
The molecular biology you need to understand the following can be found in the 5 posts in the “Molecular Biology Survival Guide”. Here’s the link https://luysii.wordpress.com/category/molecular-biology-survival-guide/
The parts of the genome we understand the best are the parts that code for protein. The Cancer Genome Atlas spent a lot of money (1.5 Billion $) looking at genes coding for protein. They started with the most highly curated (translation: hopefully valid) set of genes. There have been tons of papers coming out of the project. Unfortunately, all we know is that every cancer studied so far has lots of mutations in the protein coding genes (the average breast or colon tumor has 90+ such mutations). Many of the mutations found were in previously known cancer causing genes (oncogenes). It would be great if, say, the same few genes were mutated in a given type of cancer. Unfortunately this isn’t the case. There is very little overlap in the mutations between two tumors of the same type. Of course this bears out the well recognized clinical fact not all tumors from organ X (say breast) act the same, even when they are sub classified by the way they look under the microscope (histological type).
This brings us to [ Science vol. 339 pp. 957 – 959 – 961 ’13 ]. Once you venture away from the parts of of the genome coding for the amino acids of a protein, the next best understood parts, are those immediately in front (5′ to) the gene itself. The closest element is called the promoter, and is usually within 100 base pairs of the site at which transcription of DNA into RNA begins. Then farther out (sometimes thousands of base pairs out) come the enhancers, which help transcription factors do their work.
TERT is a protein which helps to keep the ends of chromosomes (telomeres) intact. This work looked at the promoter rather than the protein itself, and found two mutations. Together they were found in an astounding 50/70 melanomas, and 24/150 of a variety of cancer cell lines. I think this is much, much higher than any particular oncogene was found mutated in the Cancer Genome Atlas. The net effect of the two mutations was to make binding of a type of transcription factor (ETS) easier, resulting in more TERT being made.
So it’s time to start looking at the 98% of the genome NOT coding for protein. The problem with looking here, is that we really don’t know what it’s doing. The days when it was called junk are mercifully behind us. The second problem, is that we’re going to find changes from the ‘standard’ genome (which really doesn’t exist — the average infant contains 30 mutations not present in either parent) and we have no clue as to how to interpret them.