How machines are dictating the methods of curing AIDS?
The HIV virus has been responsible for 32 million deaths across the globe and the number keeps growing. Although the medical community has made massive contributions to finding a cure to this deadly virus, some of the current methods of treatment that are under development include: A first class medicine treatment which works by preventing the human immunodeficiency virus from breaking through the cell membrane, a cell therapy method that works by modifying the patient’s own cells to become resistant to the HIV virus and methods like attachment inhibitor, gene modification, involving T- cell response etc.
The infection contains highly active antiretroviral therapy or also known as HAART. But the anti-retroviral drugs are expensive and thus creates a need for finding methods and medicines which improves the life expectancy of the patients without showing any harmful side effects. This is where our reliance on machine learning starts increasing as we turn to them to help us find cost less curing methods reachable to a common man.
Machine learning algorithms such as pattern recognition, regression, classification and prediction are extensively used in such fields of research. While classification algorithms are implemented to tell between active and inactive compounds, we use the available data to train the regression algorithms. The ensemble algorithms that include bagging and boosting are then used to make faster and accurate decisions. Machine learning algorithms are also being used in quantitative structure activity relationship and ligand based virtual screening. Modern QSAR(Quantitative Structure Activity Relationship) is based on chemical structures developed with extensive use of both linear and non-linear optimization techniques while also having a strong emphasis on model validation. Support Vector Machines, Ensemble Methods, Naive Bayes Classifiers, Neural Networks are few of the many methods being pursued to find a solution to this virus and in drug development.
The spectrum of applications of these methods extends over a massive area. Although researchers acknowledge the difficulty to find a cure for the virus due to its high mutability, they are still dedicating huge efforts to find a way around. These techniques do simplify the solution but are still complicated to draw inference from and implement them. In the end, algorithms and lines of code continue to make this world a better place.