The paper featured on the cover of this issue of Aging-US, published on October 30, 2025, entitled “SAMP-Score: a morphology-based machine learning classification method for screening pro-senescence compounds in p16-positive cancer cells,” represents an important methodological and conceptual advance at the interface of senescence biology, imaging and drug discovery.

In this study, led by first author Ryan Wallis and corresponding author Cleo L. Bishop (Queen Mary University of London), the authors introduce SAMP-Score, a machine-learning–based framework designed to identify bona fide senescence induction in cancer cells where canonical markers fail. This is a timely and much-needed contribution to the field.

Therapy-induced senescence has emerged as a powerful strategy to restrain tumor growth, yet its reliable detection in cancer cells remains a major bottleneckIn these contexts, cells often already display features associated with cellular aging, rendering conventional senescence markers ambiguous or misleading. Distinguishing true senescence from toxicity, stress responses or baseline “aged” phenotypes is therefore a critical unmet need.

Rather than relying on predefined molecular readouts, the authors take a different approach and train a machine-learning model to recognize senescence-associated morphological profiles (SAMPs) which are subtle but reproducible changes in cellular architecture captured through high-content microscopy. By learning directly from image-based phenotypes, SAMP-Score is able to identify senescence with a level of precision that is difficult to achieve using marker-based strategies alone.

The strength of the platform demonstrated through a large-scale screen of over 10,000 novel chemical entities in p16-positive basal-like breast cancer cells. From this screen, the compound QM5928 emerged as a robust inducer of senescence across multiple cancer models, notably without inducing cytotoxicity. Importantly, QM5928 retains activity in cellular contexts that are resistant to CDK4/6 inhibition, including palbociclib-refractory, p16-high tumors.

Mechanistically, the authors show that QM5928 promotes nuclear relocalization of p16, consistent with a functional engagement of cell-cycle arrest pathways. These nuanced phenotypic changes would likely have gone undetected without the resolution and discrimination provided by SAMP-Score, underscoring the platform’s ability to separate true senescence from confounding cellular states.

This work exemplifies how machine learning and quantitative imaging can be harnessed to solve long-standing problems in senescence research, moving the field beyond binary marker expression toward phenotype-driven classification. Beyond its immediate relevance for cancer therapy, SAMP-Score offers a broadly applicable framework for senescence-based screening efforts across biological contexts.

DOI - https://doi.org/10.18632/aging.206333

Corresponding author - Cleo L. Bishop - [email protected]

Abstract video - https://www.youtube.com/watch?v=qXI_KI3EgHE

Sign up for free Altmetric alerts about this article -

https://aging.altmetric.com/details/email_updates?id=10.18632%2Faging.206333

Subscribe for free publication alerts from Aging - https://www.aging-us.com/subscribe-to-toc-alerts

To learn more about the journal, please visit https://www.Aging-US.com​​ and connect with us on social media at:

Bluesky - https://bsky.app/profile/aging-us.bsky.social

ResearchGate - https://www.researchgate.net/journal/Aging-1945-4589

Facebook - https://www.facebook.com/AgingUS/

X - https://twitter.com/AgingJrnl

Instagram - https://www.instagram.com/agingjrnl/

YouTube - https://www.youtube.com/@Aging-US

LinkedIn - https://www.linkedin.com/company/aging/

Pinterest - https://www.pinterest.com/AgingUS/

Spotify - https://open.spotify.com/show/1X4HQQgegjReaf6Mozn6Mc