How a yeast cell helps crack open the 'black box' behind artificial intelligence

March 5, 2018, University of California - San Diego
How a yeast cell helps crack open the 'black box' behind artificial intelligence
Screenshot from d-cell.ucsd.edu, where researchers can use DCell, a new virtual yeast cell developed at UC San Diego School of Medicine. Credit: UC San Diego Health

UC San Diego School of Medicine researchers developed a visible neural network and used it to build DCell, a virtual model of a functioning brewer's yeast cell.

"It seems like every time you turn around, someone is talking about the importance of artificial intelligence and machine learning," said Trey Ideker, PhD, University of California San Diego School of Medicine and Moores Cancer Center professor. "But all of these systems are so-called 'black boxes.' They can be very predictive, but we don't actually know all that much about how they work."

Ideker gives an example: machine learning systems can analyze the online behaviors of millions of people to flag an individual as a potential "terrorist" or "suicide risk." "Yet we have no idea how the machine reached that conclusion," he said.

For machine learning to be useful and trustworthy in health care, Ideker said, practitioners need to open up the and understand how a system arrives at a decision.

Machine learning systems are built on layers of artificial neurons, known as a neural network. The layers are tied together by seemingly random connections between neurons. The systems "learn" by fine-tuning those connections.

Ideker's research team recently developed what they call a "visible" neural network and used it to build DCell, a model of a functioning brewer's yeast cell, commonly used as a model in basic research. To do this, they amassed all knowledge of cell biology in one place and created a hierarchy of these . Then they mapped standard machine learning algorithms to this knowledgebase.

DCell can be viewed at d-cell.ucsd.edu. The technical details are published on March 5 in Nature Methods.

But what excites Ideker the most is that DCell is not a black box; the connections are not a mystery and cannot form by happenstance. Instead, "learning" is guided only by real-world cellular behaviors and constraints coded from approximately 2,500 known cellular components. The team inputs information about genes and genetic mutation and DCell predicts cellular behaviors, such as growth. They trained DCell on several million genotypes and found that the virtual cell could simulate cellular growth nearly as accurately a real cell grown in a laboratory.

"Human knowledge is incomplete," said Jianzhu Ma, PhD, an assistant research scientist in Ideker's lab who led the efforts to build DCell. "We want to complete that knowledge to help guide predictions, in and elsewhere."

Ideker and Ma also put DCell to the test. If they deliberately fed the system false information, it wouldn't work. Take ribosomes, for example. Cells use these tiny biological machines to translate genetic information into proteins. But if the researchers instead wired ribosomes to an unrelated process like apoptosis, a system use to commit suicide, DCell could no longer predict cell growth. The virtual cell "knows" that the new arrangement isn't biologically possible.

Ideker and his colleagues at the Cancer Cell Map Initiative, which he co-directs, are now generating some of the experimental data they need to build a DCell for human cancer. Then they will determine how best to personalize this virtual cell approach for a patient's unique biology.

"We want one day to be able to input your specific cancer-related genetic mutations and get back a readout on how aggressive your cancer is, and the best therapeutic approach to prevent its growth and metastasis," said Ideker, who is also founder of the UC San Diego Center for Computational Biology and Bioinformatics.

Explore further: Researchers ID cancer gene-drug combinations ripe for precision medicine

More information: Using deep learning to model the hierarchical structure and function of a cell, Nature Methods (2018). nature.com/articles/doi:10.1038/nmeth.4627

Related Stories

'Wildly heterogeneous genes'

September 15, 2013

Cancer tumors almost never share the exact same genetic mutations, a fact that has confounded scientific efforts to better categorize cancer types and develop more targeted, effective treatments.

Gene editing technique helps find cancer's weak spots

March 20, 2017

Genetic mutations that cause cancer also weaken cancer cells, creating an opportunity for researchers to develop drugs that will selectively kill them, while sparing normal cells. This concept is called "synthetic lethality" ...

Recommended for you

Pushing lithium ion batteries to the next performance level

December 13, 2018

Conventional lithium ion batteries, such as those widely used in smartphones and notebooks, have reached performance limits. Materials chemist Freddy Kleitz from the Faculty of Chemistry of the University of Vienna and international ...

Uber filed paperwork for IPO: report

December 8, 2018

Ride-share company Uber quietly filed paperwork this week for its initial public offering, the Wall Street Journal reported late Friday.

1 comment

Adjust slider to filter visible comments by rank

Display comments: newest first

Eikka
not rated yet Mar 05, 2018
For machine learning to be useful and trustworthy in health care, Ideker said, practitioners need to open up the black box and understand how a system arrives at a decision.


There's the fundamental paradox. If you understood how it reaches the correct conclusion, you would be capable of making that conclusion yourself and you wouldn't need the AI because you could just code the same thing into a simple dumb "IF x THEN y" algorithm that checks a few boxes and gives you the answer. Artifical intelligence becomes artifical stupidity.

As long as you don't understand the conclusion, you're in no position to evaluate whether it is actually correct or merely lucky, and therefore you can't trust your AI.

This is also why AIs can't develop other smarter AIs. If the new AI is any more intelligent it's rather by accident than by design.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.