iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
No—not a ready-to-use treatment. In a 2023 University of Toronto project, researchers used AI tools to design and synthesize a potential hit molecule aimed at hepatocellular carcinoma (HCC), the most common primary liver-cancer type. The first effort took 30 days from target selection and involved synthesizing seven compounds. The result was an early laboratory discovery, not a drug tested in patients or approved for use.
What did researchers accomplish in 30 days?
The University of Toronto team combined predicted protein-structure information from AlphaFold with the Pharma.AI platform. Its PandaOmics tool supported biocomputation, while Chemistry42 was used for generative chemistry. The researchers reported identifying a previously undiscovered HCC target and pathway, then designing a “novel hit molecule” intended to bind that target without relying on an experimentally determined protein structure.
The reported 30 days ran from target selection to the first hit molecule. The team synthesized seven compounds in that initial effort. A second design round produced a more potent hit, but potency in this discovery work did not establish that the molecule was safe, effective, or suitable for treating people.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Alán Aspuru-Guzik, a University of Toronto professor of chemistry and computer science, described the broader aim this way: “If one uses a generative model targeting an AI-derived protein, one can substantially expand the range of diseases that we can target.”
#1 Best Overall
- New
- Mint Condition
- Dispatch same day for order received before 12 noon
- Guaranteed packaging
- No quibbles returns
What does “potential hit molecule” mean?
A hit is an early candidate that shows enough promise against a biological target to justify further investigation. It is a starting point for drug discovery, not the same thing as a validated medicine. A more potent hit may perform better in a particular laboratory assay, but that alone says little about whether it can reach the right tissues, work at a tolerable dose, or improve health outcomes.
The Toronto report described a potential drug and said clinical trials were still needed. It did not report human dosing, safety, efficacy, regulatory approval, or patient outcomes for the 30-day molecule. The project therefore did not complete drug development in a month.
Is the molecule available to patients, or tested in humans?
The cited Toronto report does not establish that the molecule was given to people or made available as a treatment. It reports discovery-stage work and says clinical trials remained necessary. Without human testing and the evidence required for regulatory review, patients and clinicians cannot treat it as an established therapy.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsHow is AI being used elsewhere in liver-related research?
AI work in liver disease spans different tasks. Screening possible drug candidates, measuring biopsy features in trials, and helping detect cancer are not interchangeable with inventing a treatment or demonstrating that one improves patient outcomes.
Rank #3
| Example | What the AI is used for | What it does—and does not—show |
|---|---|---|
| FDA SmartCore project | AI-driven screening of primary tumor tissue for candidate identification or repurposing in fibrolamellar carcinoma. | The project plans validation in patient-derived xenograft models. It is a research platform, not an approved therapy. |
| FDA-qualified AIM-NASH, December 8, 2025 | Helping pathologists score MASH liver-biopsy features in clinical trials. | It supports trial measurement, not cancer treatment. Pathologists still review the full slide and accept or reject AI scores. |
| LiON study in Nature Medicine | AI-assisted liver-cancer detection. | It evaluates diagnosis, not therapy; strong diagnostic performance does not by itself establish improved clinical outcomes. |
What the LiON results can—and cannot—tell us
The LiON study trained its system on 6,443 patients, validated it across 22,251, and tested it as an additional reader in 10,333 routine-care patients. In its single-arm trial, it reported an area under the curve (AUC) of 0.952, with a 95% confidence interval of 0.942–0.961. The authors said prospective comparative studies across diverse systems are still needed. These findings concern diagnostic performance; they do not show that AI created a treatment or that its use improves survival.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Can AI replace oncologists or clinical trials?
No evidence in the Toronto report supports that conclusion. AI helped researchers generate a candidate, but discovery is only an early part of establishing a treatment. Clinical trials are needed to assess safety and whether a candidate benefits people, and regulatory review remains part of the path to approval.
AI recommendations also do not automatically change clinical decisions. A National Cancer Institute summary published February 21, 2025, reported mixed clinician acceptance of AI treatment recommendations and greater reluctance to change liver-cancer decisions when an AI recommendation departed from standard care. Issam El Naqa, MD, a senior author cited in the summary, cautioned: “Today’s AI tools aren’t perfectly accurate, and can be biased and limited depending on the quality of their training data.”
Recommended Free Tools
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
