The age written in a passport says little about what is happening inside an individual cell. Two cells from the same organism can exist in completely different biological states: one retains signs of youth, while another already exhibits stress, DNA damage, and the characteristic features of cellular senescence.
Researchers from the Karolinska Institutet and Stockholm University have developed a tool called Pasta that allows this difference to be seen through gene activity. The work was published on 27 July 2026 in the journal Advanced Science. Pasta belongs to so-called transcriptomic aging clocks—systems that estimate biological age based on which genes in a cell are active and how strongly they are functioning.
Age as a state
To create Pasta, the researchers collected data on the gene activity of 17 212 healthy tissue samples from 21 studies. The model incorporated data from different tissues and various gene analysis technologies.
Many previous transcriptomic clocks work well only with a specific tissue or a particular measurement technology. The researchers attempted to create a model capable of recognizing a general age-related shift regardless of where the sample was obtained.
To do this, they used an age-shift approach. Instead of simply trying to guess a person's chronological age, the algorithm learned to recognize which of two comparable samples displays a younger or older gene expression profile.
As a result, Pasta was able to work with bulk RNA-seq, single-cell RNA-seq, and microarray data.
The resulting indicator is therefore more accurately called a cell's transcriptomic biological age rather than its absolute "true age."
What happens to an aging cell?
When the researchers looked at which biological processes were most strongly associated with Pasta's readings, a familiar pattern of aging emerged.
High transcriptomic age was associated with p53 pathways, cellular stress, and responses to DNA damage. The model also distinguished between younger, stem-like cell states and the state of cellular senescence—when a cell stops dividing normally and shifts into a stably altered functional mode. The researchers went further. They decided to use the clock not only to measure aging but also to search for what is capable of shifting it.
Three million experiments through the eyes of one algorithm
Pasta was applied to the massive Connectivity Map L1000 database, which contains over three million gene activity profiles. In these experiments, cells were exposed to medicinal substances or genetic changes, after which researchers measured how gene function was rearranged.
The analysis included 248 cell lines, more than 30 000 chemical treatments, and about 14 000 genetic interventions. For each treatment, Pasta asked essentially one question: did the cell's transcriptomic profile look older or younger?
In this way, the researchers discovered 271 compounds associated with a shift toward an older state and 63 compounds associated with a shift toward a younger transcriptomic profile. This is precisely where Pasta transforms from a "clock" into a tool for searching for potential mechanisms of aging.
Unexpected mechanisms of aging
Disruptions in mitochondrial translation processes—the production of proteins inside mitochondria—were found to be associated with states that shift cells toward aging. Conversely, mRNA splicing—a process in which a cell edits a primary RNA molecule before creating a protein—was linked to a shift toward a younger state.
This is particularly interesting because aging here appears not as the work of a single "aging gene," but as a change in an entire network of processes: DNA repair, mitochondrial function, RNA processing, stress responses, and gene expression regulation.
A cell can be older than its organism
The Pasta study adds another argument in favor of an idea that is increasingly changing the biology of aging: age is not just the time that has passed since birth. It is a state of the system. Within a single organism, different cells and tissues can move along the trajectory of aging at different speeds. And while age was previously measured primarily by the calendar, today there is an increasing attempt to read it directly in the molecular architecture of the cell—in the epigenome, proteins, metabolites, and now, with increasing precision, in the transcriptome.
Pasta is not yet a clinical test and cannot tell a person how many years they have left to live. But it offers researchers something else: a fast way to check whichinterventions shift the molecular state of a cell toward aging, and which ones shift it in the opposite direction.
And this is precisely where the important significance of the work lies. If biological age can be measured as a changeable state, the next question arises almost inevitably: how far can this state actually be turned back?
