New Study Debunks One Gene Myth of Scientists, but Others Persist

Ever since DNA was discovered in the middle of the twentieth century, many scientists have tended to exaggerate and overstate its importance. Ignoring the physical limitations of DNA, which put the most severe limit on what information can be expressed in DNA, many scientists have trumpeted DNA as some “secret of life” or “blueprint” for constructing humans. Many scientists have spoken as if DNA is some “master molecule” that determines our health, our lifespans, our physical shape and even our mental characteristics. Actual observations have repeatedly clashed and conflicted with such exaggerated ideas.

The first big disappointment for those hoping to prove “genes are our destiny” came in the early twenty-first century, after the completion of the Human Genome Project. As this long project proceeded over multiple years, those who exaggerated the role of DNA repeatedly told us that once the Human Genome Project was finished analyzing and mapping all of the genes in human DNA, there would be some tremendous benefit to the public. We were told again and again by the experts that the Human Genome Project would lead to a host of medical advances that would greatly impact the average person. In 2000 genetic research leader Francis Collins said, “Over the longer term, perhaps in another 15 or 20 years, you will see a complete transformation of therapeutic medicine.”

No such thing has happened. The medical advancements produced by the Human Genome Project have been slight. Speaking of the Human Genome Project, a 2010 article in the New York Times is entitled, “A Decade Later, Genetic Map Yields Few New Cures.”

In 2011 Jonathan Latham stated the following in The Guardian:

The failure to find meaningful inherited genetic predispositions is likely to become the most profound crisis that science has faced. Not only has the most expensive scientific project ever conceived failed to reach a goal it assured the world it would achieve, but there is also the ticklish problem of why the headlines have been so consistently discrepant with reality. As the failures to find significant genes have accumulated, geneticists have remained silent.

A recent scientific study exposes how far off scientists have been in their exaggerations about genes. The study is about how much genes determine a person's lifespan. Based on all the hype we have heard about DNA being “the secret of life” and “the master molecule,” you would think that perhaps 60% or 70% of the variation in lifespans between different people can be explained by their genes. Scientists had previously estimated that only between 15% to 30% of the variation in lifespans between people can be explained by genes.

But the new study tells us that such estimates have been way off – not because they are too low, but because they are too high. The study estimates that “well less than 10%” of the variation in lifespans between different people can be explained by their genes. Table 1 of the paper gives estimates of only about 5% for how much the variation in lifespans between different people can be explained by their genes.

So apparently previous scientists have overestimated by between 300% and 600% the degree to which genes affect lifespans (as estimating a 5% effect as a 30% effect is an overestimation of 600%). Are there any other “gene whoppers” our scientists have been telling us? There are a few of these.

One is the frequently stated idea that instincts or behavior tendencies might be the result of genes. It is easy to see why that cannot be true. DNA basically lists only the chemical ingredients of proteins. Essentially the only thing you can state in DNA are which amino acids make up a protein. There is no way in which DNA could be expressing anything like behavior rules. There is no way in which anything like a rule such as “be afraid of snakes” could be expressed in DNA.

Another erroneous idea frequently advanced by biologists is the idea that DNA is a blueprint for constructing an organism. There are a number of powerful reasons why this idea must be false, which I explain in detail here and here. The most powerful reason is that DNA uses a minimalist language in which basically the only words are the names of 20 amino acids. Such a language is sufficient for expressing the linear sequence of amino acids that make up a polypeptide chain, but completely insufficient for expressing complex three-dimensional information such as the shape and structure of an eye or the shape and structure of a reproductive system. Another powerful reason why DNA (and the genes it is made of) cannot be a blueprint for a human body plan is that human DNA has been thoroughly analyzed, through huge scientific studies such as the Human Genome Project and the ENCODE project, and we have found no evidence that human body plan information is stored in it. Also, we know that some species with vastly simpler structures than humans have many more genes than humans. The simple rice plant has thousands of more genes than humans, the opposite of what would be true if genes stored body plans. The total size of human DNA is megabytes is much smaller than the total size of the DNA in megabytes of many plants. 

The information in DNA is one-dimensional information, and one-dimensional information is completely insufficient to specify the three-dimensional information needed to specify a body plan.  The total size of the human genome is only about 700 megabytes, much less than the genome size of a tomato plant, and way too small to store a specification of a human.  A modern CT scanner uses three times more megabytes just to store images of a body that don't even include functional information.  

The set of observable characteristics in an organism is called its phenotype. The information in an organism's DNA is called its genotype. Phenotype is not predictable from genotype. Although genotypes affect phenotypes, genotypes do not specify phenotypes or body plans. Similarly, your educational level affects but does not specify your annual income. In the mainstream book "Frontiers in Ecology, Evolution and Complexity," a scientist states the following:

At the beginning of the 21st century, biology confronted an uncomfortable fact: despite the increasing availability of whole genome sequence data, it was not possible to predict, or even clarify, phenotypic observations. In fact, we now know that there is not sufficient information in the linear DNA of the complete genomes to recover and/or understand the diverse phenotypic states of an organism. 

Massimo Pigliucci (mainstream author of numerous scientific papers on evolution) has stated that "old-fashioned metaphors like genetic blueprint and genetic programme are not only woefully inadequate but positively misleading."

Why does there persist the myth that body plans are specified in DNA? It's because this is part of Neo-Darwinian dogma. The prevailing version of Darwinism (called Neo-Darwinism or “the modern synthesis”) makes or assumes the following claims:

1. Body plans and phenotypes are specified by DNA.
   
2. Mutations in DNA cause one species to evolve into another species.
   

The first claim is false. Body plans and phenotypes are not specified by DNA. It is therefore impossible that we could ever explain the evolution of one species into another species merely by assuming DNA mutations, regardless of what effect natural selection had on such mutations. It is conceivable that there is some undiscovered mechanism of information storage or hereditary that might allow one species to evolve into another, but the idea that this occurs merely because of changes in DNA cannot be correct.

We can imagine a device that scientists would have invented by now if it were true that genomes specify phenotypes or body plans. The device (which could be a computer program or collection of computer programs) is one that we can call a general phenotype prediction machine. The device would take as an input the DNA data of a particular organism. The device would then be able to predict what type of macroscopic organism that DNA corresponds to. So if you gave the device the DNA data of a mouse, the device would be able to describe a mouse-like organism. And if you gave the device the DNA data of an elephant, the device would be able to predict an elephant-like organism.

But an important rule of this device is that it could not simply do a pattern match using a database of stored genomes. So the device could not simply check its database to determine which of its many stored genomes was the closest match to the genome supplied as an input. Instead, the device would actually have to deduce the characteristics of the organism based on information in the input DNA. So, for example, if the device were to predict that the organism had two arms, it would have to find some information in the input DNA that specified or implied such a conclusion.

This phenotype prediction device would have to list the characteristics that it predicted that the organism had, and also justify each characteristic that it listed. So if the device declared that the predicted organism had two eyes rather than six eyes or no eyes, it would have to explain what it found in the input DNA to justify such a conclusion. And if the device declared that the predicted organism had two wings rather than zero wings, the device would have to explain what it found in the input DNA that specified such a thing.

Nothing like the phenotype prediction device I have described has ever been created, at least not one that works. The closest thing I can find to the phenotype prediction device I described is one in a 2017 paper with the misleading title “DeepMetabolism: A Deep Learning System to Predict Phenotype from Genome Sequencing.” But the system works using multiple types of biological databases (violating my rules), and only allows you to tell which of three types of E coli microbes some input DNA comes from. That's light-years from being able to predict the phenotype of a large organism from only its genotype. The misleadingly-titled paper should have been entitled, “DeepMetabolism: A Deep Learning System to Predict E Coli Microbe Phenotypes from Genome Sequencing and Lots of Other Data.”

No such general phentoype prediction device like I have described will ever be invented and successfully used, unless scientists cheat by sneaking in pattern matching using a database of stored genomes. This is because genotypes do not specify phenotypes. Or to say pretty much the same idea, DNA does not specify body plans. Neither DNA nor genes nor genomes contain a blueprint or set of instructions for building an organism. 

Shockingly, we do not at all know where phenotypes or body plans come from. You will sometimes hear people say that phenotype is caused by a combination of genotype and "the environment," but this is misleading. There is nothing in the environment of a newly fertilized ovum that might cause it to turn into a human rather than some very different biological form.  An honest and candid statement about the source of phenotypes or body plans would be to say that they must come mainly from some very important facet of reality that we do not understand. You can call it the gigantic missing link of biological life. 

But since scientists do not like to admit that their biological knowledge is only fragmentary, you will continue to hear lots of “DNA is a body blueprint” falsehoods in the years ahead. We have been grossly misinformed about DNA and genes for a long, long time, and such misrepresentations will not soon end.