By Sophia Wolfram
This is an edited version of an essay submitted to the Newnham College Biological Sciences Essay Prize. If you would like to share your essay competition entry, please contact us.
In the last 50 years we have made more progress in the battle to maintain human life than in all preceding history. This advancement has been almost all due to an explosion in the understanding of biology. We have increased life expectancy, lowered infant mortality and even eradicated some deadly diseases. These are phenomenal achievements; however they are the result of a change much more fundamental than scientific research alone, which if recognised and embraced correctly, will open the floodgates to medical discovery.
So what is this driving power? I believe the answer is computation.
Computation has changed nearly every corner of the modern world. It will revolutionise medicine, but it may also force it on its head. Historically, the treatment of disease has been approached from the outside in; beginning by treating the symptoms after an illness has reared its head. Though our methods of doing this have changed (drastically for the better) and we have treatments which can attack the root cause of illness, we are still playing medical catch-up—especially with chronic diseases. To ensure our success rate increases, we have to embrace preventative (and in many cases, personalised) medicine.
Preventative medicine is harder to set up than acute treatment but it can save many more lives, by stopping the problem before it has begun. This seems fairly obvious, however it is much easier said than done. To predict which people might get which illness, we need to understand a multitude of factors which have a strong link to each condition. While this doesn’t sound very challenging, once the different factors are explored, the complexity multiplies quickly. The factors could be genetic, environmental, lifestyle or other sporadic medical conditions. Each of these factors gives us thousands of data points which add up to give the total risk of developing a chronic disease. Of course, to have a benchmark of the risk we also need a reference dataset. Some datasets already exist in a crude form—it is how we know that smoking causes lung cancer, or that living in poverty can increase heart disease risk 3-6 fold. The problem is that they are not detailed enough for accurate prediction, which is where computation comes in.
We need as many data points as possible on each person to build a holistic model of them. This can be through routine tracking (for weight, travel history, smoking etc.) or—our best bet for prediction—through genetic sequencing. Genetic sequencing is an incredible tool for prediction as it literally uncovers the instructions for our body, including those for disease. Computation allowed us to make the leap from the chemical composition inside each cell to the abstract sequence we know. There are nearly 3 billion base pairs in a human genome. All the instructions needed for a fully functioning human being reside in the sequence of 4 bases, a sequence that runs 3 billion base pairs long. That is an astonishing number, it seems both too small to produce the great variation we see in the human race, and simultaneously much too large to store, even for a potential diagnosis. For now, it may be out of reach, but we won’t have to wait much longer.
The first finalised human genome was published in 2003 at the end of a long eight year project to decipher the human code. We have reduced the cost from more than a billion dollars to just 3000 and reduced the time from years to just 1 or 2 days, all thanks to technology and computation. It is astonishing that in just 17 years (my lifetime!) it could be drastically transformed. Looking at this, it is far easier to imagine a world where genome sequencing is standard at birth or the first port of call in diagnosis or discussing risk of a disease. If that world of medicine comes true, we will have shifted to personalised medicine.
Personalised medicine is the future of medical care. Facilitated by improving computational power, it can give us the crystal ball medicine needs, and prevent the pain of many chronic illnesses. Admittedly it cannot do everything. It cannot predict if someone will be in a car crash or if a toddler eats something they shouldn’t, so we still need the acute treatment we have now. However, the two should work in partnership so that acute treatment is used for acute illnesses, not trying to patch up chronic conditions that cannot be reversed.
Preventative medicine will save more lives. This is clearly the future we want, so we need to make it happen.
References:
Comments