Brian Clarke
@brianfclarke
Group Leader at German Cancer Research Center (DKFZ). Moving to BlueSky, please join me there: https://bsky.app/profile/brianfclarke.bsky.social
I'll be shutting down my (admittedly very quiet) account on Twitter/X soon, so please follow me on BlueSky: bsky.app/profile/brianf…
#GPMelt users, it was fantastic meeting so many of you at my #HUPO2024 poster!🎊 As discussed, the revised GPMelt code is now on GitLab. Also, check out the GPMelt website to dive into the model—no advanced maths required! 💻🔍 #Proteomics All links here: linktr.ee/CecileCADS
Herzlichen Dank an die Stabstelle für Öffentlichkeitsarbeit des DKFZ für diese allgemeinverständliche Pressemitteilung über unsere Arbeit
Forschende aus dem DKFZ, dem @embl und von der @TU_Muenchen haben einen auf Deep Learning basierenden Algorithmus vorgestellt, der die Auswirkungen seltener Erbgut-Varianten auf die Gesundheit präziser vorhersagen kann. ➡️ t1p.de/9x3f0
Big thanks to the DKFZ PR department for this writeup of our results suitable for a general audience
Researchers from DKFZ, @embl and @TU_Muenchen have introduced an algorithm based on deep learning that can predict the effects of rare genetic variants on health more precisely. @NatureGenet t1p.de/5zbo8
🧵1/10 We’re thrilled to share our latest paper, Bicycle, a novel approach for inferring cyclic causal relationships from observational and interventional data! This work addresses a critical gap in understanding complex biological networks.
📢 PhD/Postdoc opportunities in our group: - tinyurl.com/ml-spatial-data - tinyurl.com/grn-multi-omics - tinyurl.com/causal-single-… Join our vibrant research community in Heidelberg if you are excited about research at the interface of statistics, machine learning and life sciences! Plz RT
Wanna do Rare Variant Association Testing powered by deep learning? See @brianfclarke & my poster PB4113 Fr. 3pm on DeepRVAT. New results on phenome-wide application, binary traits, 10x faster than SOTA @OliverStegle & @gagneurlab collab. Update: shorturl.at/egtB0 #ASHG23
1/n We're very excited to announce the release of #DeepRVAT, a deep neural network approach to learn burden scores from rare variants by integrating dozens of annotations in a data-driven manner.
What a joy it's been to work on this with @Holtkamp_Eva @OliverStegle @gagneurlab and the rest of the team at DKFZ, TUM, and @AIHCluster Very proud of these advances in using deep learning to model rare variants in association genetics and prediction of high-risk individuals
1/n We're very excited to announce the release of #DeepRVAT, a deep neural network approach to learn burden scores from rare variants by integrating dozens of annotations in a data-driven manner.