As we’ve seen before, AI and machine learning has started to become a majorly used technology in the medical industry. By equipping computers with the ability to almost mimic human intelligence, mankind has been able to use AI as an aid in several scenarios. In some cases AI has even help us achieve feats that are not achievable through regular human effort.
For instance, AI and machine learning algorithms have helped biologists make sense of the vast number of molecular signals that control how genes function. But, as new algorithms are developed to handle even larger amounts of data, the complexity behind interpreting the results of such operations also increases. For this reason, to help biologists understand results easier, quantitative biologists, Justine B. Kenney and Ammar Tareen have come up with a strategy to design better, more interpretable machine learning algorithms.
The scientists created these efficient new algorithms by taking inspiration from the structure of a human brain. Kinney and Tareen’s algorithm was designed to work in a way similar to how neurons connect and branch. This neural network that they created comes under a category of algorithms termed ANN ( artificial neural network). ANNs are said to be the basis for the next generation of advanced machine learning algorithms as the potential applications of these type of networks are massive.
By using the power of these ANNs, Kinney and Tareen examine data through an experimental method called massively parallel reporter assay. Then, by using this information, quantitative biologists can ANNs predict which molecules control which specific genes during gene regulation.
Although Kinney and Tareen’s method, as stated above is a type of ANN, they offer an additional feature that sets it apart from currently available options. Today, most standard ANNs are shaped from MPRA data which is a lot different from how scientists ask questions in real life. This makes it difficult for biologists to interpret the gene regulation data.
But with the new approach that our duo have developed, scientists may be able to understand how proteins turn and off to regulate genes in a much clearer manner. Their method essentially bridges the gap between how biologists think and the computational tools that are involved. Kinney and Tareen’s ANN was designed in such a way that it mathematically reflects common concepts in biology concerning genes and the molecules that regulate them. By doing so, the method can basically process and display data that is more understandable for a biologist.
Comprehending and monitoring the tiniest of biological responses involved in gene regulation can be a challenging task. And when the readability of related data is difficult, it can make the process a whole lot more complex than it already is. But now with this new type of ANN, biologists may just be able to get the break they deserve. Moreover, with this method being extended to cover a wide variety of biological systems, it could quite possibly create a revolution within the medical space.