iiinova — where your digital self keeps evolving

Machine Learning in Biomedicine @mlbiomed.bsky.social · 04/09/2026

Yrjö Koski presented his work at IEEE CIBCB 2026 titled NanoDS: Accurate simulation of nanopore sequencing data with distribution approximations 🧬💻 cibcb2026.dib.uth.gr

011

Reposted by Machine Learning in Biomedicine

Esa Pitkänen @epitkanen.bsky.social · 16/06/2026

Really happy to see this one out! Great work by @lalangohr.bsky.social, Ilse Kaaja and the whole team. 🩸🧬 @mlbiomed.bsky.social @fimm-uh.bsky.social

012

Reposted by Machine Learning in Biomedicine

Esa Pitkänen @epitkanen.bsky.social · 01/09/2026

Can domain-specific foundation models compete with large pan-tissue models? Yes. HistoEncoder, trained on prostate tissue, matches H-Optimus-1, UNI2-h, Virchow2 and CONCH v1.5 in cancer detection and grading, while being much smaller and 3–5× faster. doi.org/10.1016/j.jp...

111

Reposted by Machine Learning in Biomedicine

Esa Pitkänen @epitkanen.bsky.social · 06/11/2025

Excited to share our latest pre-print on profiling DNA adducts with nanopore sequencing! This study was led by Yrjö Koski, who developed a computational toolkit called IonStats, and identified several compound-specific effects caused by genotoxic compounds. www.biorxiv.org/content/10.1...

biorxiv.org

Compound-specific DNA adduct profiling with nanopore sequencing and IonStats

Covalently bound DNA adducts are mutation precursors that contribute to aging and diseases such as cancer. Accurate detection of adducts in the genome will shed light on tumorigenesis. Commonly used a...

322