iiinova — where your digital self keeps evolving

Jérôme Salignon @jeromesalignon.bsky.social · 09/06/2025

Can we measure & shift a cell’s age? Meet Pasta, a plug-and-play transcriptomic clock that flags perturbations that ↑ (cancer therapy) or ↓ (regenerative medicine) biological age. doi.org/pqhh bsky.app/profile/jero...

doi.org

011

Jérôme Salignon @jeromesalignon.bsky.social · 05/06/2025

🧬Integration with DepMap highlighted the pivotal role of epigenetic regulation and splicing pathways in shaping cellular responses to pro-rejuvenating signals, and of mitochondrial translation and type I interferon in responses to pro-aging signals. #RegenerativeMedicine

100

Jérôme Salignon @jeromesalignon.bsky.social · 05/06/2025

💊 FDA-approved anti-cancer drugs were highly enriched in Pasta’s pro-aging hits—strongly supporting the idea that chemotherapy works by aging cancer cells. This data-driven approach may help uncover safe, cancer-specific therapies. #Chemotherapy

Horizontal bar chart: anti-cancer drugs enriched among pro-ageing hits; AML bar tallest, most other cancers show similar trend.

100

Jérôme Salignon @jeromesalignon.bsky.social · 05/06/2025

💻 Screening 3 million CMAP profiles revealed pro-aging and pro-rejuvenation genetic and chemical perturbations. Importantly, these two sets were enriched in known pro-senescence and pro-stemness factors, respectively.

Twin bar charts: ageing compounds (olive bars) enriched for senescence; rejuvenating compounds (navy) enriched for reprogramming.

100

Jérôme Salignon @jeromesalignon.bsky.social · 05/06/2025

🩺In TCGA tumors, Pasta’s scores correlated better with tumor grade than chronological age, indicating its value as a biological age predictor. Pasta’s scores were also predictive of patient survival for certain cancers. #CancerResearch

100

Jérôme Salignon @jeromesalignon.bsky.social · 05/06/2025

🔍A novel age-shift modeling strategy yielded more accurate and biologically meaningful predictions, based on p53-related genes, than traditional regression models. Pasta also cleanly separated senescence, quiescence, stemness & differentiation in static and dynamic settings.

Two scatter plots: replicative senescence shows age score rising with day; OSKM reprogramming shows score dropping over time.

100

Jérôme Salignon @jeromesalignon.bsky.social · 05/06/2025

📊 Trained on 21 heterogeneous datasets to predict relative age, Pasta was designed to be broadly applicable to many kinds of transcriptomic data (i.e., microarrays, L1000…). It outperformed other RNA-Seq-specific tools without requiring reprocessing or batch correction.

Dot plot: red (Pasta) vs purple (MultiTIMER) Pearson r values across 14 datasets—Pasta higher in 11/14.

100

Jérôme Salignon @jeromesalignon.bsky.social · 05/06/2025

131

Reposted by Jérôme Salignon

Floris Barthel @florisbarthel.bsky.social · 16/04/2025

🚨 Pre-print alert! 🚨 We generated chromosome-scale diploid genome assemblies for the widely used fibroblast lines BJ and IMR-90. These models of replicative senescence and telomere shortening now have fully resolved telomeres, subtelomeres, and centromeres. 🧬 🧵👇

195