The Sydney Morning Herald
· collected 2026-10-11 · by Sally Rawsthorne
Professor Fatemeh Vafaee at the University of New South Wales in Sydney is using artificial intelligence to revolutionize cancer research and drug development. Her lab has developed an AI platform that predicts effective drug combinations before lab testing, significantly speeding up the process compared to traditional methods, which have a 10% success rate over 15 years and cost about $2 billion per drug. This technology has already identified 24 synergistic drug combinations for cancer treatment, with some progressing to clinical trials. Vafaee’s work also includes developing a blood test to rule out recurrent breast cancer, affecting 15% of patients. The University of NSW will highlight these advancements in its upcoming Social Impact of Science report.
Written locally by qwen2.5:14b on 2026-10-11,
using this article's own text rather than the other coverage of the
same event (that is the story summary below).
Story summary
Professor Fatemeh Vafaee, deputy director of the University of NSW’s AI institute, is leveraging artificial intelligence to improve cancer treatment outcomes. Her lab in Sydney aims to develop personalized medications, including treatments for rare cancers, through advanced AI drug-development platforms. Traditional drug development has a success rate of about 10% and takes around 15 years, costing approximately $2 billion per drug. Vafaee's approach uses AI to identify effective combinations of drugs more efficiently than previous methods, potentially accelerating the discovery process and improving patient outcomes.
Written for “AI In Cancer Research” on 2026-10-11,
grounded in this article and the 0 other(s) covering the same event.
Why this leaning score
This article does not take a side on a contested political
question, so it has no leaning score. That is an
answer rather than a gap: a match report or a rescue can be warmly
or critically written without being left or right, and scoring it
anyway is how approval of a subject gets recorded as a political
position.
No political leaning scored for article 71752 · logged 2026-10-11
Personalised medication, drugs to combat the rarest of cancers, knowledge of our individual cell interactions and even pre-empting disease in the body – one Sydney lab is working on all of these possibilities and more, thanks to the advent of artificial intelligence.
asserted
lab → combat → intelligence
“I’ve become very interested in how we can leverage AI to save lives and improve the patient outcomes,” said Professor Fatemeh Vafaee, deputy director of the University of NSW’s AI institute and a professor in the school of biotechnology and biomolecular sciences.
asserted
Vafaee → become → biotechnology
An AI drug-development platform is among the breakthroughs made possible since the technology exploded in 2022, Vafaee said.
asserted
Vafaee → make → 2022
The model is upending traditional drug development, which succeeds around 10 per cent of the time, takes 15 years to develop a drug from beginning to end and costs about $2 billion.
asserted
which → upend → billion
The complexity of diseases such as cancer and diabetes means that often one medication is insufficient and combination therapy – using a mix of different drugs together – is better able to tackle the disease.
asserted
therapy → mean → disease
But the number of potential combinations had previously slowed researchers down.
asserted
number → slow → researchers
“It’s an intractable problem – if you have 1000 drugs and you want to look into the combination of two or three drugs across different dosages, there are millions of experiments that you need to run,” said Vafaee.
asserted
Vafaee → ’ → that
“This is where AI can come on board to prioritise that and look into combinations that could be efficacious before we go into the lab experiments.
uncertain
we → come → experiments
“We can predict possible combinations of drugs ahead of lab experiments.”
asserted
We → predict → experiments
Vafaee’s lab has so far picked 24 synergistic combinations, which have been validated in pre-clinical studies at Melbourne’s Peter MacCallum Cancer Centre and are now going to higher levels of testing.
asserted
which → pick → testing
Earlier this year, Vafaee worked with a commercial partner to develop a blood test ruling out recurrent breast cancer, something that affects 15 per cent of breast cancer patients and that had plagued doctors because of its unpredictable post-cancer time frame.
asserted
that → work → frame
These breakthroughs are among the work highlighted by the University of NSW in its forthcoming Social Impact of Science report, which will be released on Monday.
asserted
which → highlight → Monday
The report proposes a new framework for measuring research that zeroes in on commercialisation, scholarly outputs, policy and influence, lives changed, and sustainability and development.
asserted
that → propose → commercialisation
Born in Iran, Vafaee received her PhD in AI in 2011 from Chicago’s University of Illinois, which she has since applied to biomedicine.
asserted
she → bear → biomedicine
AI developments and access to large-scale data since 2022 – the same year that ChatGPT launched – have meant that scientists can do things “several thousand times faster” than previously.
asserted
scientists → launch → things
“We don’t want just to predict for breast cancer and brain cancer,” she said.
asserted
she → want → cancer
“We want to be able to adopt that into [rare] cancer or even other diseases.
asserted
We → want → cancer
“And this is where we have been developing AI tools for transfer learning, what we can transfer the knowledge from one disease to another ... that would help us to go through less common diseases and look at how we adopt AI for the patient’s benefits.
asserted
we → develop → benefits
Every single patient matters, and we can use this technology for patients that we have less population level data available for.”
asserted
data → matter → that
Personalised medicine for all was around the corner, she said, if AI was granted access to large-scale data.
asserted
AI → say → data
“We are all different and we need to move on from one solution fit everyone,” she said.
asserted
she → need → everyone
“What I would love to see in future is that we go from being reactive to what we observe, to being proactive to what we predict.”
asserted
we → love → what
The Aporia
