CytoData-2019

Europe/Berlin
Lecture Hall (German Cancer Research Center / Communication Center)

Lecture Hall

German Cancer Research Center / Communication Center

Im Neuenheimer Feld 280 69120 Heidelberg Germany
Florian Heigwer (Deutsches Krebsforschungszentrum), Christian Scheeder (German Cancer research Center (DKFZ)), Benedikt Rauscher (German Cancer Research Center (DKFZ)), Michael Boutros (German Cancer Research Center (DKFZ) )
Description

Attendees will join together for a one-day symposium with seminars, lectures and social events followed by a two-day hackathon. We aspire to gather a diverse and engaged group of both image-based profiling experts, and members of related fields. Everyone involved will be an active contributor (even if remote), no passive absorption of others’ ideas allowed! We assume participants will be willing to share their analysis strategies, given that each laboratory’s “competitive advantage” in this field does not come from their data pre-processing pipeline but rather from computational techniques downstream of the steps covered in the event, and from biological discoveries derived from their unique data.

Sun Oct 27 Educational session (optional)
Learn about the latest tools and methods in single cell and image based profiling.

Mon Oct 28 Full day Symposium
Hear from leading researchers in the field from academia and industry.

Tue Oct 29 CytoData Hackathon - Day 1
Get oriented, and starting working on the first stage of the challenge.

Wed Oct 30 CytoData Hackathon - Day 2
Refine models using second stage of the challenge, then evaluate on held out set.
Winners will be announced based on the held out set.

Contact Field for Persons or Groups
    • 14:00 15:15
      Tutorials: From deep learning to Interactive Vizualization: Tutorials: Deep learning Buchleithner room Ground floor (TP3)

      Buchleithner room Ground floor

      TP3

      Im Neuenheimer Feld 580 69120 Heidelberg
      • 14:00
        Tutorials: High-throughput Image Analysis with Deep Learning 1h 15m Buchleithner Seminar room

        Buchleithner Seminar room

        TP3

        Im Neuenheimer Feld 580

        Deep learning has emerged as a powerful tool to analyze imaging data, with many successful applications in computer vision. It is increasingly being adopted in biological experiments involving high-throughput imaging, to identify single cells, classify hits, explore cellular phenotypes and generate digital markers. In this workshop, we aim to give an overview of what deep learning is, what are the fundamental concepts, and the main applications that researchers have explored in high-throughput imaging experiments.

        Content and Schedule:

        14:00 Part I - Introduction (​by Oren Kraus​) - 20 minutes
        ● Intro to neural networks (math of neurons + gradient descent)
        ● What happened around 2010? bigger datasets + GPU + algorithms
        ● Models: Alexnet -> object recognition, U-Net -> medical image segmentation
        ● Hands on: TensorFlow Playground - ​https://playground.tensorflow.org/

        14:20 Part II - Applications (​by Oren Kraus and Juan Caicedo​) - 30 minutes
        ● Workflow 1) classification (single cell or whole field)
        ● Workflow 2) Image repurposing
        ● Workflow 3) Profiling (single cell, weakly supervised, autoencoders)
        ● Workflow 4) Cell segmentation
        ● Workflow 5) In silico labeling - predict markers/probe

        14:50 Part III - Discussions (​by Juan Caicedo)​ - 10 minutes
        ● Concerns & future directions
        ● Hands-on or live discussion.

        Speakers: Juan Caicedo (Broad Institute), Oren Kraus (Phenomic AI)
    • 14:00 15:15
      Tutorials: From learning on images to open science: Machine learning on images Glassbox 3.OG (TP3)

      Glassbox 3.OG

      TP3

      Im Neuenheimer Feld 580 69120 Heidelberg

      Workshops

      • 14:00
        ilastik: a simple and user-friendly machine learning toolkit for image analysis 1h 15m
        Speaker: Anna Kreshuk (EMBL, Heidelberg)
    • 15:15 15:30
      Break/Change of rooms 15m TP3

      TP3

    • 15:30 16:45
      Tutorials: From deep learning to Interactive Vizualization: Interactive Visualization Buchleithner room Ground floor (TP3)

      Buchleithner room Ground floor

      TP3

      Im Neuenheimer Feld 580 69120 Heidelberg
      • 15:30
        Single-cell RNA-Seq data analysis for pedestrians 1h 15m

        There are many workflows and packages to analyse single-cell RNA-Seq data, which all automate a great deal of the process. While this is convenient, it makes it hard to understand what is going on "under the hood" and to adapt and modify specific steps to one's needs, or to notice peculiarities in intermediate results. In this workshop, we will perform a typical RNA-Seq analysis using (nearly) only basic R functions in order to explore how standard workflows really work (and discover that there is much less to it than many practitioners may think), and see how to "hack" them to get more from your data.

        Speaker: Simon Anders (ZMBH-Heidelberg University)
    • 15:30 16:45
      Tutorials: From learning on images to open science: Open Science forum Glassbox 3.OG (TP3)

      Glassbox 3.OG

      TP3

      Im Neuenheimer Feld 580 69120 Heidelberg

      Workshops

      • 15:30
        Open Science Forum: Principles and Practice 1h 15m

        Open science is a method of conducting research in which all software, data, results, experiments, and interpretations are made publicly available upon project initiation. In this forum, we will discuss open science principles and best practices, with a focus on maintaining open science lab notebooks. Are you interested in increasing exposure to your science, improving public trust in your results, and increasing the speed of scientific discoveries? If so, come learn about and discuss open science best practices. Initiate open science projects and don’t do invisible work!

        Speaker: Gregory Way (Broad Institute)
    • 09:00 10:25
      Deciphering biology by functional genomics and microscopy-based profiling Lecture Hall

      Lecture Hall

      German Cancer Research Center / Communication Center

      Im Neuenheimer Feld 280 69120 Heidelberg Germany
      • 09:00
        Welcome 10m
        Speaker: Florian Heigwer (Deutsches Krebsforschungszentrum)
      • 09:10
        Seeing is believing – using imaging to map gene-gene and gene-drug interactions at scale 45m
        Speaker: Michael Boutros (German Cancer Research Center (DKFZ) )
      • 09:55
        A computational framework to build a dynamic protein atlas of human cell division 30m

        Cell division is a fundamental process of life where a mother cell undergoes profound morphological transformations to divide into two and passes its genetic information to the daughter cells. This process is driven by a tight coordination of a large number of essential proteins in both space and time. Recent publication from our group (Cai, Hossain et. al., Nature, 2018) reported an integrated experimental and computational approach for profiling essential mitotic proteins on a single cell level. A dynamic protein atlas was constructed where users can visualize distributions of mitotic proteins during cell division and extract various parameters such subcellular localization of proteins in different stages of mitosis, absolute number proteins and average concentration inside different cellular compartments. These parameters revealed many biological insights to understand cell division in a comprehensive manner. This talk will mainly focus on the computational aspects of the project in terms of standardizing the single cell data in both space and time to demonstrate the power of computational approaches to answer fundamental biological questions.

        Speaker: Hossain Julius (EMBL)
    • 10:25 10:45
      Coffee Break 20m Lecture Hall

      Lecture Hall

      German Cancer Research Center / Communication Center

      Im Neuenheimer Feld 280 69120 Heidelberg Germany
    • 10:45 12:15
      Deciphering biology by functional genomics and microscopy-based profiling Lecture Hall

      Lecture Hall

      German Cancer Research Center / Communication Center

      Im Neuenheimer Feld 280 69120 Heidelberg Germany
      • 10:45
        Imaging-based single cell biology at large scale 45m

        High throughput microscopy-based approaches combined with automated genome-scale systematic perturbations of single genes have generated a wealth of high-quality data demonstrating the potential of coming a step closer to a comprehensive functional annotation of the human genome and the underlying gene-gene interactions. Interestingly, detailed analyses of the frequency of occurrence of phenotypes in such datasets in relation to the success rate of the perturbations applied, show a significant variability of the occurrence of abundant phenotypes despite a successful know-down/knock-out of the respective target genes, supporting the hypothesis that single cells have or are able to develop molecular compensation mechanisms in response to specific gene knock-downs or knock-outs. In order to identify such molecular compensation mechanisms underlying the regulation of the secretory pathway we have developed an automated imaging-based workflow which allows automated recognition of Golgi phenotypes in response to gene knock-downs followed by photoactivatable fluorescent marking of phenotypic cells in order to sort them subsequently by flow cytometry and analyse their transcriptome at the single cell or phenotype level. First results using this workflow support the existence of robust molecular compensation pathways related to cell signaling, cell survival or regulation of the early secretory pathway.

        Speaker: Rainer Pepperkok (EMBL)
      • 11:30
        The Influence of the Bone Marrow Niche on Drug Response Phenotypes of Blood Cancers 30m
        Speaker: Sophie Rabe (University Hospital Heidelberg)
    • 12:15 13:15
      Lunch Break 1h Lecture Hall

      Lecture Hall

      German Cancer Research Center / Communication Center

      Im Neuenheimer Feld 280 69120 Heidelberg Germany
    • 13:15 15:00
      Emerging imaging technologies and analyses to empower personalized medicine approaches Lecture Hall

      Lecture Hall

      German Cancer Research Center / Communication Center

      Im Neuenheimer Feld 280 69120 Heidelberg Germany
      • 13:15
        Single cells in a whole organism: combining morphology and gene expression 45m
        Speaker: Anna Kreshuk (EMBL)
      • 14:00
        Spatial single-cell metabolomics in situ identifies cell states and subpopulations 30m

        Growing appreciation of the key roles of metabolism and recent revelations of the extent of
        cellular heterogeneity demand to investigate metabolism in single cells. We developed SpaceM,
        a method for in situ spatial single-cell metabolomics of adherent cells that detects >100
        metabolites in >10.000 individual cells together with fluorescent markers and morpho-spatial
        cellular features. We validated SpaceM by predicting the cell type of co-cultured human cells
        and mouse fibroblasts, revealing metabolic intermixing between neighboring cells. Differentiated
        human hepatocytes with progressed steatosis induced by fatty acids and proinflammatory
        cytokines showed accumulation of specific lipid classes e.g. longer phospholipids and shorter
        neutral lipids. We characterized the metabolic state of steatotic hepatocytes and metabolic
        plasticity associated with NF-κB. We discovered that steatosis and proliferation take place in
        distinct cell subpopulations, each with a characteristic spatial organization and metabolic
        signatures. SpaceM enables spatio-molecular investigations of metabolomes in single cells in
        various cellular models.

        Speaker: Luca Rappez (EMBL)
      • 14:30
        Implementation of ex vivo high-content functional drug response profiling in paediatric precision oncology 30m
        Speaker: Sina Oppermann (German Cancer Research Center (DKFZ))
    • 15:00 15:20
      Coffee Break 20m Lecture Hall

      Lecture Hall

      German Cancer Research Center / Communication Center

      Im Neuenheimer Feld 280 69120 Heidelberg Germany
    • 15:20 18:05
      From association to function: Identification and validation of drug targets in single cells Lecture Hall

      Lecture Hall

      German Cancer Research Center / Communication Center

      Im Neuenheimer Feld 280 69120 Heidelberg Germany
      • 15:20
        Quantification and Modeling of Single Blood Cells 55m

        Time-lapse microscopy is a powerful method to continuously monitor single cells. Combined with appropriate markers, it allows for the time-resolved observation of protein expression. Such information is necessary to characterize state transitions, differentiation dynamics and gene expression models. I will present bioimage informatics tools, computational models and deep learning approaches to infer cellular properties from single-cell data.

        In lineage trees of differentiating blood stem cells, stem cell decisions happen before an observable state change, inducing correlations in sister cells. Using these correlations and a stochastic model for a differentiation process, we find differentiation events to happen much earlier than anticipated and identify transcription factor expression that is inconsistent with the current toggle switch paradigm. To predict differentiation prospectively, we use a deep neural network trained on image patches from brightfield microscopy and cellular movement. Based on these features, we can detect lineage choice in blood stem cells up to three generations before conventional molecular lineage markers are observable. Our deep learning approach is also useful for clinical applications: Using digitized images of single healthy and leukemic cells from clinical blood smears we are able to discriminate diagnostic blast cell reaching expert performance.

        The combination of statistical and mechanistic models opens up possibilities for precisely quantifying single-cell dynamics and developing new diagnostic approaches.

        Speaker: Carsten Marr (Helmholtz Zentrum München Institute of Computational Biology)
      • 16:15
        Single cell genomics 45m
        Speaker: Oliver Stegle (EMBL)
      • 17:00
        Defining best practices for automated outlier detection from single cell images 15m
        Speaker: Nil Sahin (University of Toronto)
      • 17:15
        Multi-Template Matching: a versatile tool for object-localization in microscopy images 15m
        Speaker: Laurent Thomas (ACQUIFER is a division of DITABIS AG)
      • 17:30
        Quantifying co-cultures with weakly supervised and multiple instance learning 15m
        Speaker: Oren Kraus (Phenomic AI)
      • 17:45
        Robust morphological profiling of immunological synapses in natural killer lymphocytes 15m
        Speaker: Loan Vulliard (CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences)
      • 18:00
        Closing Remarks 5m
        Speaker: Florian Heigwer (Deutsches Krebsforschungszentrum)
    • 19:30 21:30
      Speakers Dinner (Invited speakers only): Salerno's, Mönchhofstrasse Heidelberg Salernos

      Salernos

    • 09:00 10:00
      Introduction Seminar room K2 (DKFZ Communication center)

      Seminar room K2

      DKFZ Communication center

      Im Neuenheimer Feld 280 69120 Heidelberg
      • 09:00
        Formulating the Challenge 30m
        Speaker: Florian Heigwer (Deutsches Krebsforschungszentrum)
      • 09:30
        Explaining the data 30m
        Speaker: Christian Scheeder (German Cancer research Center (DKFZ))
    • 10:00 12:00
      Familiarize with the data: Hacking seminar room K2 (DKFZ Communication center)

      seminar room K2

      DKFZ Communication center

      Im neuenheimer Feld 280 69120 Heidelberg
    • 12:00 13:00
      Lunch Break 1h Lecture Hall

      Lecture Hall

      German Cancer Research Center / Communication Center

      Im Neuenheimer Feld 280 69120 Heidelberg Germany
    • 13:00 18:00
      Hackathon: Hacking seminar room K2 (DKFZ Communcation center)

      seminar room K2

      DKFZ Communcation center

      Im neuenheimer Feld 280 69120 Heidelberg
      • 13:00
        Hacking 2h
        Speaker: ALL
      • 15:00
        Coffee Break 30m
      • 15:30
        Hacking 2h 30m
        Speaker: ALL
    • 09:00 12:00
      Hackathon: Hacking Semianr room K2 (DKFZ Communcation Center)

      Semianr room K2

      DKFZ Communcation Center

      Im neuenheimer Feld 280 69120 Heidelberg
    • 12:00 13:00
      Lunch Break 1h Lecture Hall

      Lecture Hall

      German Cancer Research Center / Communication Center

      Im Neuenheimer Feld 280 69120 Heidelberg Germany
    • 13:00 16:00
      Hackathon: Hacking Seminar room K2 (DKFZ Communication Center)

      Seminar room K2

      DKFZ Communication Center

      Im neuenheimer Feld 280 69120 Heidelberg
    • 16:00 18:00
      Final: Wrapping up Seminar room K2 (DKFZ Communication Center)

      Seminar room K2

      DKFZ Communication Center

      Im neuenheimer Feld 280 69120 Heidelberg
      • 16:00
        Coffee Break 30m
      • 16:30
        Final pitches of individual solutions 30m
        Speaker: ALL
      • 17:00
        Award Ceremony 30m
        Speaker: Florian Heigwer (Deutsches Krebsforschungszentrum)