BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:DKTK School of Oncology Tech Talk - Putze
DTSTART:20260923T140000Z
DTEND:20260923T150000Z
DTSTAMP:20260826T074800Z
UID:indico-event-1682@indico.dkfz.de
CONTACT:dktk-schoolofoncology@dkfz-heidelberg.de\;+49 6221 42-1681
DESCRIPTION:DKTK School of Oncology Tech Talk \nwith Philipp Putze\n \nP
 hD candidate at the Technical University of Munich (TUM) and the German Ca
 ncer Research Center (DKFZ)\nWednesday 23.09.26 | 16:00 - 17:00 via Zoom\n
 Join our second DKTK School of Oncology Tech Talk in 2026 with Philipp Put
 ze who will talk about "From Noise to Insight: A Computational Journey Thr
 ough Single-Cell Biology". After the presentation there will be ample oppo
 rtunity to ask questions to the speaker and to engage in a lively discussi
 on.\n \nSingle-cell and spatial transcriptomics have transformed our abil
 ity to study biology\, scaling from experiments with hundreds of cells to 
 atlases containing millions. Yet this explosion in data has brought equall
 y significant computational challenges\, including technical noise\, batch
  effects\, and the integration of increasingly complex modalities. In this
  talk\, I will trace the evolution of computational methods that have enab
 led this progress\; from linear approaches such as PCA\, which first helpe
 d to reveal cellular diversity\, to deep generative models that disentangl
 e biological signal from technical variation while enabling the prediction
  of rich molecular measurements from more abundant and cost-effective assa
 ys. Finally\, I will discuss the emergence of single-cell foundation model
 s\, which move beyond reconstructing noisy observations to learning transf
 erable representations of biology itself\, allowing new datasets to be int
 erpreted in the context of millions of previously measured cells and openi
 ng new opportunities for cross-dataset integration\, multimodal patient st
 ratification\, and biological discovery.\n \n\n \nSPEAKER:\nPhilipp Putz
 e\n\nPhilipp Putze is a PhD candidate in the lab of Dieter Saur (DKTK Prof
 essorship) at the Technical University of Munich (TUM) and the German Canc
 er Research Center (DKFZ)\, where he develops deep learning methods for si
 ngle-cell and spatial transcriptomics. His research combines machine learn
 ing with cancer biology to study tumor ecosystems and develop computationa
 l tools that enable biological discovery at atlas scale. His work spans mu
 ltimodal transcriptional analysis\, generative modeling\, and the developm
 ent of foundation models for single-cell biology. Before joining TUM\, he 
 completed his Master’s degree in Multidisciplinary Research in Experimen
 tal Sciences through a joint project at the Insitute for Research in Biome
 dicine Barcelona (IRB) and the Barcelona Supercomputing Center (BSC)\, whe
 re he investigated the early evolution of pancreatic cancer using single-c
 ell transcriptomics.\n \n\nVENUE:\nThe Talk will take place online via Zo
 om.\nRegistered participants will receive the login information.\n\nORGANI
 ZER & HOST:\nDKTK School of Oncology for DKFZ\n\n \n \n\n \n\nhttps://i
 ndico.dkfz.de/event/1682/
IMAGE;VALUE=URI:https://indico.dkfz.de/event/1682/logo-286316080.png
LOCATION:Zoom online
URL:https://indico.dkfz.de/event/1682/
END:VEVENT
END:VCALENDAR
