The Discovery of a Breakthrough Antibiotic Using Artificial Intelligence

With the rise of bacteria that are resistant to drug treatment, it is predicted that 10 million people could die by the year 2050. Should one be so unlucky to be afflicted with any of these antibiotic-resistant bacteria, then there’s no hope unless a breakthrough is discovered to fight against known pathogens that pose a threat to the human race.

The utilization of AI has proven to be effective in the discovery of a new and potent antibiotic. This technological advancement presents a unique way of uncovering a breakthrough for the world at large. It has also paid off for leading businesses looking to take advantage of cutting-edge technology fully.

With the help of a group of MIT researchers, artificial intelligence technology has revealed a potent molecule with the ability to kill a good number of life-threatening pathogens. Compared to the long-standing antibiotic ciprofloxacin, bacteria like E. Coli failed to stand up in resistance to Halicin.

As decades have passed and we’ve seen the effectiveness of technology (artificial intelligence) in bringing out a potent drug to fight against antibiotic-resistant bacteria. In time past, predictive computers and AI have not proven to yield the best results. But with new AI technology models, the world could be hosting a breakthrough for the pharmaceutical sector, as well as for business leaders.

Presently, molecules are easily mapped as continuous vectors and picked up on representation by the latest neural networks to predict the properties of the drug.

With new rising conditions affiliated with antibiotics resistant pathogens, there’s an evergreen demand tied to breakthrough antibiotics discovery. The conventional methods for the discovery of antibiotics can be daunting and time-consuming. But with the new AI innovation, this stress can be reduced by a sizable factor and the time cut down to a few weeks.

Halicin Discovery

Before discovering a potent antibiotic molecule, over 6,000 molecules of the Drug Repurposing Hub were initially screened, which made it a tremendous challenge rediscovering the exact compound of interest.

To make the work easier, compounds that exhibit a different composition and structure from existing antibiotics were only picked for further tests. After cycles of study, the AI system was finally able to pick up on one molecule with high potency against Escherichia coli (E. Coli).

Halicin further went under more traditional laboratory tests to uncover how powerful it is at fighting against various bacteria species and know its mechanism in doing so.

It was later discovered that Halicin meddles with the bacteria’s ability to successfully sustain a means of generating Adenosine Triphosphate (ATP), which is the energy unit for every living cell. This causes the bacteria to die on exposure to Halicin.

Application of Machine Learning

Within a short frame of time, machine learning can effectively screen over millions of chemical compounds. This drug discovery model is being developed and implemented to enhance the selection of potential and active antibiotics that can eliminate bacteria through several methods of known antibiotic drugs.

Research teams aim to develop an enabling platform that gives people the privilege of leveraging machine learning power. The same approach has helped researchers isolate and identify the most potent molecule.

The machine learning algorithm per se is not new, being that we’ve always carried around the popular example in our pockets.

It started with the researchers building a database (DeepARG-DB) of recognized resistance genes and a range of 30 drugs affected. This database was obtained from the Comprehensive Antibiotic Resistance Database, the Universal Protein Resource, and the Antibiotic Resistance Genes Database to help facilitate positive results.

The researchers later trained the machine learning algorithm using only 70 percent of the entire 10,602 resistance genes from the Universal Protein Resource. Coming up with the data imputed, researchers have made comparisons between the individual sequence of the genes collected to the recognized gene resistant to drugs from the two other databases.

As a result, coming out was an array of thousand genes having stark similarity to each gene collected from the Universal Protein Resource.

The Zhang’s Group Research Using Machine Learning

As every other research team has done to uncover the potency locked up in a micro molecule, Zhang’s group also came up with a critical learning technique.

The team’s machine-learning algorithm came as an inspiration on how the brain of human functions. They achieved the technique by imputing tons of ideas into the process to ensure the best result is recorded.

In the course of the research, the AI could track down weights of similar genes, accurately predicting the category of antibiotic resistance. Two models of the machine were built to carry out effective tests on different DNA sequences.

What AI Technology Holds For The Future

The use of machine learning algorithms can revolutionize the conventional, time-consuming technique of discovering antibiotics.

The Broad Institute reported on how the discovery of antibiotics using machine learning could be a breakthrough to science and the rest of the world.

The effect of this discovery is enormous for different areas of life to effectively fight against drug-resistant pathogens.

Researchers have also stated that machine learning has also been just the start of something new for the world, looking at how it has displayed itself to be a piece of unique and powerful equipment.

There could be an improvement in artificial intelligence to enhance the screening of a larger library of bacteria to discover new molecules that exhibit antimicrobial activities in the future.

More Opportunities In-Store

AI’s possibilities in tailoring in-depth machine learning algorithms that will nail potent molecules and fight actively against special types of drug resistance genes are endless.

Antibiotic-resistant bacteria are thought to be common among soldiers because of recruits’ exposure to various medical environments. This can lead to an increased number of deaths among soldiers.

For instance, Halicin is highly potent against A. baumannii – a bacteria resistant to all recognized antibiotics (presently a big problem among the United States soldiers being recruited in Afghanistan).

With more investment made on developing more efficient AI, there’ll be a better chance of fighting against life-threatening pathogens.

In the future, the research team looks forward to the intravenous delivery of newly developed antibiotics to the exact site that is infected so that there can be perfect elimination without meddling with harmless bacteria in the human body.

Conclusion

As a result of the fast-growing economy of the fight against drug-resistant pathogens, researchers have taken it in hand to learn machine algorithms to help identify genes that are resistant to drugs in a pathogen.

A good number has deployed artificial intelligence (AI) to track down new resistant genes by seeking to understand how these bacteria dust off drug treatments.

However, some seek to understand more about how the resistance profile of certain environments (like wastewater) is. This has raised the need for researchers to move into metagenomes.