Tag Archives: Larisa N. Soldatova

Transformational machine learning (TML)

Leave a reply

It seems machine learning is getting a tune-up. A November 29, 2021 news item on ScienceDaily describes research into improving machine learning from an international team of researchers,

Researchers have developed a new approach to machine learning that ‘learns how to learn’ and out-performs current machine learning methods for drug design, which in turn could accelerate the search for new disease treatments.

The method, called transformational machine learning (TML), was developed by a team from the UK, Sweden, India and Netherlands. It learns from multiple problems and improves performance while it learns.

…

A November 29, 2021 University of Cambridge press release (also on EurekAlert), which originated the news item, describes the potential this new technique may have on drug discovery and more,

TML could accelerate the identification and production of new drugs by improving the machine learning systems which are used to identify them. The results are reported in the Proceedings of the National Academy of Sciences.

Most types of machine learning (ML) use labelled examples, and these examples are almost always represented in the computer using intrinsic features, such as the colour or shape of an object. The computer then forms general rules that relate the features to the labels.

“It’s sort of like teaching a child to identify different animals: this is a rabbit, this is a donkey and so on,” said Professor Ross King from Cambridge’s Department of Chemical Engineering and Biotechnology, who led the research. “If you teach a machine learning algorithm what a rabbit looks like, it will be able to tell whether an animal is or isn’t a rabbit. This is the way that most machine learning works – it deals with problems one at a time.”

However, this is not the way that human learning works: instead of dealing with a single issue at a time, we get better at learning because we have learned things in the past.

“To develop TML, we applied this approach to machine learning, and developed a system that learns information from previous problems it has encountered in order to better learn new problems,” said King, who is also a Fellow at The Alan Turing Institute. “Where a typical ML system has to start from scratch when learning to identify a new type of animal – say a kitten – TML can use the similarity to existing animals: kittens are cute like rabbits, but don’t have long ears like rabbits and donkeys. This makes TML a much more powerful approach to machine learning.”

The researchers demonstrated the effectiveness of their idea on thousands of problems from across science and engineering. They say it shows particular promise in the area of drug discovery, where this approach speeds up the process by checking what other ML models say about a particular molecule. A typical ML approach will search for drug molecules of a particular shape, for example. TML instead uses the connection of the drugs to other drug discovery problems.

“I was surprised how well it works – better than anything else we know for drug design,” said King. “It’s better at choosing drugs than humans are – and without the best science, we won’t get the best results.”

Here’s a link to and a citation for the paper,

Transformational machine learning: Learning how to learn from many related scientific problems by Ivan Olier, Oghenejokpeme I. Orhobor, Tirtharaj Dash, Andy M. Davis, Larisa N. Soldatova, Joaquin Vanschoren, and Ross D. King. PNAS December 7, 2021 118 (49) e2108013118; DOI: https://doi.org/10.1073/pnas.2108013118

This paper appears to be open access.

‘Eve’ (robot/artificial intelligence) searches for new drugs

Leave a reply

Following on today’s (Feb. 5, 2015) earlier post, The future of work during the age of robots and artificial intelligence, here’s a Feb. 3, 2015 news item on ScienceDaily featuring ‘Eve’, a scientist robot,

Eve, an artificially-intelligent ‘robot scientist’ could make drug discovery faster and much cheaper, say researchers writing in the Royal Society journal Interface. The team has demonstrated the success of the approach as Eve discovered that a compound shown to have anti-cancer properties might also be used in the fight against malaria.

A Feb. 4, 2015 University of Manchester press release (also on EurekAlert but dated Feb. 3, 2015), which originated the news item, gives a brief introduction to robot scientists,

Robot scientists are a natural extension of the trend of increased involvement of automation in science. They can automatically develop and test hypotheses to explain observations, run experiments using laboratory robotics, interpret the results to amend their hypotheses, and then repeat the cycle, automating high-throughput hypothesis-led research. Robot scientists are also well suited to recording scientific knowledge: as the experiments are conceived and executed automatically by computer, it is possible to completely capture and digitally curate all aspects of the scientific process.

In 2009, Adam, a robot scientist developed by researchers at the Universities of Aberystwyth and Cambridge, became the first machine to autonomously discover new scientific knowledge. The same team has now developed Eve, based at the University of Manchester, whose purpose is to speed up the drug discovery process and make it more economical. In the study published today, they describe how the robot can help identify promising new drug candidates for malaria and neglected tropical diseases such as African sleeping sickness and Chagas’ disease.

“Neglected tropical diseases are a scourge of humanity, infecting hundreds of millions of people, and killing millions of people every year,” says Professor Ross King, from the Manchester Institute of Biotechnology at the University of Manchester. “We know what causes these diseases and that we can, in theory, attack the parasites that cause them using small molecule drugs. But the cost and speed of drug discovery and the economic return make them unattractive to the pharmaceutical industry.

“Eve exploits its artificial intelligence to learn from early successes in her screens and select compounds that have a high probability of being active against the chosen drug target. A smart screening system, based on genetically engineered yeast, is used. This allows Eve to exclude compounds that are toxic to cells and select those that block the action of the parasite protein while leaving any equivalent human protein unscathed. This reduces the costs, uncertainty, and time involved in drug screening, and has the potential to improve the lives of millions of people worldwide.”

The press release goes on to describe how ‘Eve’ works,

Eve is designed to automate early-stage drug design. First, she systematically tests each member from a large set of compounds in the standard brute-force way of conventional mass screening. The compounds are screened against assays (tests) designed to be automatically engineered, and can be generated much faster and more cheaply than the bespoke assays that are currently standard. This enables more types of assay to be applied, more efficient use of screening facilities to be made, and thereby increases the probability of a discovery within a given budget.

Eve’s robotic system is capable of screening over 10,000 compounds per day. However, while simple to automate, mass screening is still relatively slow and wasteful of resources as every compound in the library is tested. It is also unintelligent, as it makes no use of what is learnt during screening.

To improve this process, Eve selects at random a subset of the library to find compounds that pass the first assay; any ‘hits’ are re-tested multiple times to reduce the probability of false positives. Taking this set of confirmed hits, Eve uses statistics and machine learning to predict new structures that might score better against the assays. Although she currently does not have the ability to synthesise such compounds, future versions of the robot could potentially incorporate this feature.

Steve Oliver from the Cambridge Systems Biology Centre and the Department of Biochemistry at the University of Cambridge says: “Every industry now benefits from automation and science is no exception. Bringing in machine learning to make this process intelligent – rather than just a ‘brute force’ approach – could greatly speed up scientific progress and potentially reap huge rewards.”

To test the viability of the approach, the researchers developed assays targeting key molecules from parasites responsible for diseases such as malaria, Chagas’ disease and schistosomiasis and tested against these a library of approximately 1,500 clinically approved compounds. Through this, Eve showed that a compound that has previously been investigated as an anti-cancer drug inhibits a key molecule known as DHFR in the malaria parasite. Drugs that inhibit this molecule are currently routinely used to protect against malaria, and are given to over a million children; however, the emergence of strains of parasites resistant to existing drugs means that the search for new drugs is becoming increasingly more urgent.

“Despite extensive efforts, no one has been able to find a new antimalarial that targets DHFR and is able to pass clinical trials,” adds Professor Oliver. “Eve’s discovery could be even more significant than just demonstrating a new approach to drug discovery.”

Here’s a link to and a citation for the paper,

Cheaper faster drug development validated by the repositioning of drugs against neglected tropical diseases by Kevin Williams, Elizabeth Bilsland, Andrew Sparkes, Wayne Aubrey, Michael Young, Larisa N. Soldatova, Kurt De Grave, Jan Ramon, Michaela de Clare, Worachart Sirawaraporn, Stephen G. Oliver, and Ross D. King. Journal of the Royal Society Interface March 2015 Volume: 12 Issue: 104 DOI: 10.1098/rsif.2014.1289 Published 4 February 2015

This paper is open access.