Imagine the future of medicine and genetics. You wake up, look in the mirror, press a button, and a machine tells you what medicine is best for you based on your genetic makeup. The machine knows your body will react better to a certain medication and tells you to take a certain kind (or you will not heal properly). This machine knows a great deal about you based on the data it receives about your body chemistry. It has what is essentially the manual for your genetic makeup stored inside it. It can tell how you will react to certain chemicals, viruses, and environments based only on your genetic makeup. Now, this specific technology is in fact something that is possible with the advancements in medicine and the use of AI technology. With the evolution of machine learning, genomics (or the study of our genes) is changing quickly…let’s recap some of these changes.
With machine learning, large fields of data (and our over 20,000 genes) are sequenced and analyzed in a very short period of time. Humans simply can’t process all of this information like machines can. The process is slow, time consuming and susceptible to error. Yes, but an automatic sequencing machine can analyze and sequence DNA in less than a day’s time! The machine is not only cost efficient, it also knows cell biology down to DNA, RNA, and the molecular machinery of the cell. This is why it can give input into what medications are specifically useful to different individuals based on their genetics. What this also means, economically speaking, is that the AI are a huge source of revenue for the health care industry. According to predictions made by the consultant firm Frost & Sulivan, AI systems will generate about $6.7 billion in revenue by 2021. Today, companies such as Sophia Genomics, Deep Genomics, and CRISPR, are making use of AI to change the industry as we speak.
In order for the machine to recognize DNA, it is first prepared by technicians and scientists in a form that the machines can read. It is, in essence, chopped up and chemically modified. Generally, fluorescent dyes correspond to the letters T, A, G, and C. The machine, when put to work, will spit out a sequence that is known as a rough draft, this draft has errors, but eventually, it is polished into a final product without breaks or errors. When we assess genetic mutations, we begin to understand the role these mutations play in the countless numbers of diseases and disorders in existence. We can think of mutations being responsible for disorders from immune disorders to cancer to autism. With machine learning, new insights are gained into how the mutation can form anywhere in the genome and subsequently contribute to the disorder. Machine Learning is therefore a revolution in research as protein tracking algorithms recognize patterns that are made in cells.
As researchers‘ sequence and analyze DNA with the help of AI, we run into various ethical concerns. Take for instance, gene editing. If we can make changes to genes at the cellular level what are some problems that might happen in consequence? Right now, because of these ethical concerns, gene editing research is done at the private level as it is not federally funded. In addition, the FDA also prohibits embryos from being brought to term, and modified embryos are not allowed to develop beyond day one.
Despite this, scientists are still hard at work using CRISPR, a gene editing tool that runs the tests on embryos. This is done in the hopes of preventing disorders in the future. With concerns running rampant about what will happen in the future, we wonder again if science fiction is turning into a reality. All we know is that dangerous things may happen if there is misuse of the gene-editing software, and as history has revealed to us, without stringent regulatory guidelines, there are consequences as always.
Victoria Liset is strategic business & technology consultant to SMEs. She helps businesses improve their performance by using data more efficiently, and helping them to understand the implications of new technologies such as AI, Machine Learning, Big data, blockchain and IoT.