Bharath Ananthasubramaniam

@timelyscientist.bsky.social

Visiting Prof @HumboldtUni | Engineer | #Data cruncher | Rhythm lover | Improving health with #clocks | #SciComm | #TheoryInPractice | #OpenScience

The lifelong craft of scientific visualization In data-driven R&D, your model only matters if your stakeholders can understand it. 📉 Communicating insights is perhaps the only part of the process that truly "counts" at the decision-making level. But great #DataViz is harder than it looks.🧵 [1/5]

Data Science or Data Art? Building predictive models for biology and medicine is often sold as a "pure" science. 🧬 But in practice? It’s as much about art as it is about algorithms. Here is why I think we might need to rename the field to "Data Art." 🎨 [1/5]

Reading the vast majority of computational biology studies, you’d think every new method is categorically better than the last. Statistically? A highly unlikely scenario. We’ve fallen into the "Novelty Trap," where we view benchmarking as a zero-sum game. ⚖️

ML uncovers 24h rhythms in cancer 🕒🧬 Using COFE ☕, we reconstruct #circadian rhythms from human tumor biopsies across 11 #cancer types! Turns out, tumor #clocks tick in a different timezone 🕰️💊 Got data? Let’s connect! 🤝 #AIforHealth #CircadianMedicine #oncology

PLOS Biology@plosbiology.org · last yr.

Study of #circadian rhythms requires repeated sampling, but @timelyscientist.bsky.social &co present COFE, a #MachineLearning algorithm that enables virtual sampling of internal human tissues over time, revealing insights into circadian rhythms of human #tumors @plosbiology.org 🧪 plos.io/3Zx1Lt6

Workflow of Cyclic Ordering with Feature Extraction (COFE). Data without time labels are first optionally pre-filtered and then standardized. A novel cross-validation approach finds a value for the unknown sparsity parameter that guarantees good temporal ordering. Finally, COFE both assigns time labels to the samples and produces a list of rhythmic features in the data.

StuROPx bringt Studentinnen aus ganz Berlin zusammen ❤️! Danke an das #BUA und @humboldtuni.bsky.social für die Möglichkeit und Unterstützung!

Humboldt-Universität zu Berlin@humboldtuni.bsky.social · 2y ago

🥼 #WirSindBUA: Der Chronobiologe Dr. Bharath Ananthasubramaniam ist nicht nur leidenschaftlicher Forscher, sondern unterrichtet auch gern. Im Programm StuROPx der @berlinualliance.bsky.social kann er beides miteinander verbinden. Wie das funktioniert, erklärt er 👉 www.hu-berlin.de/de/pr/wir-si...

Der Chronobiologe Dr. Bharath Ananthasubramaniam redet über das Programm StuROPx