�ݺ�ߣshows by User: ospjuth

�ݺ�ߣshows by User: ospjuth / http://www.slideshare.net/images/logo.gif �ݺ�ߣshows by User: ospjuth / Thu, 08 Dec 2022 08:22:33 GMT �ݺ�ߣShare feed for �ݺ�ߣshows by User: ospjuth Automating cell-based screening with open source, robotics and AI /slideshow/automating-cellbased-screening-with-open-source-robotics-and-ai/254822533 spjuthelrig-roboautov3-221208082233-ef31cf5e
Presentation by Prof. Ola Spjuth at the ELRIG conference 'Robotics and Automation 2022' at FESTO, Esslingen, DE on 2022-12-01. ABSTRACT: We have set up an open source robotized lab for image-based drug screening and cell profiling. The lab can operate on multiple simultaneous 384-well microplates, using high-content microscopy imaging as the primary readout. Our main protocol is morphological profiling using multiplexed fluorescent dyes (Cell Painting) after cells being exposed to treatments with individual or combinations of chemical compounds, but we are increasingly using live cell imaging in brightfield for temporal analyses. In this presentation I will describe our open source efforts in automation, covering how we apply artificial intelligence / machine learning to design efficient multi-well and multi-plate experiments, and execute the generated protocols using robotics in our cell-based lab. I will also describe our approach to preprocess, filter, store and analyze the large amounts of images produced. An important application we are targeting is exploration of combination effects of drugs and environmental chemicals.]]>
Presentation by Prof. Ola Spjuth at the ELRIG conference 'Robotics and Automation 2022' at FESTO, Esslingen, DE on 2022-12-01. ABSTRACT: We have set up an open source robotized lab for image-based drug screening and cell profiling. The lab can operate on multiple simultaneous 384-well microplates, using high-content microscopy imaging as the primary readout. Our main protocol is morphological profiling using multiplexed fluorescent dyes (Cell Painting) after cells being exposed to treatments with individual or combinations of chemical compounds, but we are increasingly using live cell imaging in brightfield for temporal analyses. In this presentation I will describe our open source efforts in automation, covering how we apply artificial intelligence / machine learning to design efficient multi-well and multi-plate experiments, and execute the generated protocols using robotics in our cell-based lab. I will also describe our approach to preprocess, filter, store and analyze the large amounts of images produced. An important application we are targeting is exploration of combination effects of drugs and environmental chemicals.]]> Thu, 08 Dec 2022 08:22:33 GMT /slideshow/automating-cellbased-screening-with-open-source-robotics-and-ai/254822533 ospjuth@slideshare.net(ospjuth) Automating cell-based screening with open source, robotics and AI ospjuth Presentation by Prof. Ola Spjuth at the ELRIG conference 'Robotics and Automation 2022' at FESTO, Esslingen, DE on 2022-12-01. ABSTRACT: We have set up an open source robotized lab for image-based drug screening and cell profiling. The lab can operate on multiple simultaneous 384-well microplates, using high-content microscopy imaging as the primary readout. Our main protocol is morphological profiling using multiplexed fluorescent dyes (Cell Painting) after cells being exposed to treatments with individual or combinations of chemical compounds, but we are increasingly using live cell imaging in brightfield for temporal analyses. In this presentation I will describe our open source efforts in automation, covering how we apply artificial intelligence / machine learning to design efficient multi-well and multi-plate experiments, and execute the generated protocols using robotics in our cell-based lab. I will also describe our approach to preprocess, filter, store and analyze the large amounts of images produced. An important application we are targeting is exploration of combination effects of drugs and environmental chemicals. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/spjuthelrig-roboautov3-221208082233-ef31cf5e-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation by Prof. Ola Spjuth at the ELRIG conference 'Robotics and Automation 2022' at FESTO, Esslingen, DE on 2022-12-01. ABSTRACT: We have set up an open source robotized lab for image-based drug screening and cell profiling. The lab can operate on multiple simultaneous 384-well microplates, using high-content microscopy imaging as the primary readout. Our main protocol is morphological profiling using multiplexed fluorescent dyes (Cell Painting) after cells being exposed to treatments with individual or combinations of chemical compounds, but we are increasingly using live cell imaging in brightfield for temporal analyses. In this presentation I will describe our open source efforts in automation, covering how we apply artificial intelligence / machine learning to design efficient multi-well and multi-plate experiments, and execute the generated protocols using robotics in our cell-based lab. I will also describe our approach to preprocess, filter, store and analyze the large amounts of images produced. An important application we are targeting is exploration of combination effects of drugs and environmental chemicals.

Automating cell-based screening with open source, robotics and AI from Ola Spjuth

]]> 130 0 https://cdn.slidesharecdn.com/ss_thumbnails/spjuthelrig-roboautov3-221208082233-ef31cf5e-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Towards automated phenotypic cell profiling with high-content imaging /slideshow/towards-automated-phenotypic-cell-profiling-with-highcontent-imaging/227300820 chembioseminarki-200207191519
Presentation by Ola Spjuth (Uppsala University and Scaleout) at the Chemical Biology Seminar Series, February 6th, at Karolinska Institutet and Science for Life Laboratory, Stockholm, Sweden. ABSTRACT Phenotypic profiling of cells with high-content imaging is emerging as an important methodology with high predictive power. The true power of these methods comes when integrated into automated, robotized systems that can be run continuously and not restricted to batch analysis. One of the main challenges then becomes how to manage and continuously analyze the large amounts of data produced. In this talk I will present our efforts to establish an automated lab for cell profiling of drugs using multiplexed fluorescence imaging (Cell Painting). I will describe our computational and lab infrastructure as well as the systems, tools an methods we are developing to sustain continuous profiling of cells and continuous AI modeling. A key objective in the group is on improving screening and toxicity assessment, but also to explore predictions of mechanisms and pathways. The long-term goal is to build a closed-loop system where results from analyses are used by an AI system to design the next round of experiments and iteratively improve the confidence in predictions. Research website: https://pharmb.io]]>
Presentation by Ola Spjuth (Uppsala University and Scaleout) at the Chemical Biology Seminar Series, February 6th, at Karolinska Institutet and Science for Life Laboratory, Stockholm, Sweden. ABSTRACT Phenotypic profiling of cells with high-content imaging is emerging as an important methodology with high predictive power. The true power of these methods comes when integrated into automated, robotized systems that can be run continuously and not restricted to batch analysis. One of the main challenges then becomes how to manage and continuously analyze the large amounts of data produced. In this talk I will present our efforts to establish an automated lab for cell profiling of drugs using multiplexed fluorescence imaging (Cell Painting). I will describe our computational and lab infrastructure as well as the systems, tools an methods we are developing to sustain continuous profiling of cells and continuous AI modeling. A key objective in the group is on improving screening and toxicity assessment, but also to explore predictions of mechanisms and pathways. The long-term goal is to build a closed-loop system where results from analyses are used by an AI system to design the next round of experiments and iteratively improve the confidence in predictions. Research website: https://pharmb.io]]> Fri, 07 Feb 2020 19:15:19 GMT /slideshow/towards-automated-phenotypic-cell-profiling-with-highcontent-imaging/227300820 ospjuth@slideshare.net(ospjuth) Towards automated phenotypic cell profiling with high-content imaging ospjuth Presentation by Ola Spjuth (Uppsala University and Scaleout) at the Chemical Biology Seminar Series, February 6th, at Karolinska Institutet and Science for Life Laboratory, Stockholm, Sweden. ABSTRACT Phenotypic profiling of cells with high-content imaging is emerging as an important methodology with high predictive power. The true power of these methods comes when integrated into automated, robotized systems that can be run continuously and not restricted to batch analysis. One of the main challenges then becomes how to manage and continuously analyze the large amounts of data produced. In this talk I will present our efforts to establish an automated lab for cell profiling of drugs using multiplexed fluorescence imaging (Cell Painting). I will describe our computational and lab infrastructure as well as the systems, tools an methods we are developing to sustain continuous profiling of cells and continuous AI modeling. A key objective in the group is on improving screening and toxicity assessment, but also to explore predictions of mechanisms and pathways. The long-term goal is to build a closed-loop system where results from analyses are used by an AI system to design the next round of experiments and iteratively improve the confidence in predictions. Research website: https://pharmb.io <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/chembioseminarki-200207191519-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation by Ola Spjuth (Uppsala University and Scaleout) at the Chemical Biology Seminar Series, February 6th, at Karolinska Institutet and Science for Life Laboratory, Stockholm, Sweden. ABSTRACT Phenotypic profiling of cells with high-content imaging is emerging as an important methodology with high predictive power. The true power of these methods comes when integrated into automated, robotized systems that can be run continuously and not restricted to batch analysis. One of the main challenges then becomes how to manage and continuously analyze the large amounts of data produced. In this talk I will present our efforts to establish an automated lab for cell profiling of drugs using multiplexed fluorescence imaging (Cell Painting). I will describe our computational and lab infrastructure as well as the systems, tools an methods we are developing to sustain continuous profiling of cells and continuous AI modeling. A key objective in the group is on improving screening and toxicity assessment, but also to explore predictions of mechanisms and pathways. The long-term goal is to build a closed-loop system where results from analyses are used by an AI system to design the next round of experiments and iteratively improve the confidence in predictions. Research website: https://pharmb.io

Towards automated phenotypic cell profiling with high-content imaging from Ola Spjuth

]]> 388 0 https://cdn.slidesharecdn.com/ss_thumbnails/chembioseminarki-200207191519-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Towards Automated AI-guided Drug Discovery Labs /slideshow/towards-automated-aiguided-drug-discovery-labs/183154246 spjuthescienceacademy2019-v2-191017142124
Presentation by Ola Spjuth (Uppsala University and Scaleout Systems) on 2019-10-16 at Swedish e-Science Academy 2019 in Lund, Sweden. Research website at Uppsala University: https://pharmb.io Scaleout Systems: https://scaleoutsystems.com ]]>
Presentation by Ola Spjuth (Uppsala University and Scaleout Systems) on 2019-10-16 at Swedish e-Science Academy 2019 in Lund, Sweden. Research website at Uppsala University: https://pharmb.io Scaleout Systems: https://scaleoutsystems.com ]]> Thu, 17 Oct 2019 14:21:24 GMT /slideshow/towards-automated-aiguided-drug-discovery-labs/183154246 ospjuth@slideshare.net(ospjuth) Towards Automated AI-guided Drug Discovery Labs ospjuth Presentation by Ola Spjuth (Uppsala University and Scaleout Systems) on 2019-10-16 at Swedish e-Science Academy 2019 in Lund, Sweden. Research website at Uppsala University: https://pharmb.io Scaleout Systems: https://scaleoutsystems.com <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/spjuthescienceacademy2019-v2-191017142124-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation by Ola Spjuth (Uppsala University and Scaleout Systems) on 2019-10-16 at Swedish e-Science Academy 2019 in Lund, Sweden. Research website at Uppsala University: https://pharmb.io Scaleout Systems: https://scaleoutsystems.com

Towards Automated AI-guided Drug Discovery Labs from Ola Spjuth

]]> 391 1 https://cdn.slidesharecdn.com/ss_thumbnails/spjuthescienceacademy2019-v2-191017142124-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Combining Prediction Intervals on Multi-Source Non-Disclosed Regression Datasets /slideshow/combining-prediction-intervals-on-multisource-nondisclosed-regression-datasets/170566798 ndcp-copa2019-190910115337
Presentation given by Ola Spjuth at the 8th Symposium on Conformal and Probabilistic Prediction with Applications in Varna, Bulgaria on Sept 20th, 2019. ABSTRACT Conformal Prediction is a framework that produces prediction intervals based on the output from a machine learning algorithm. In this paper we explore the case when training data is made up of multiple parts available in different sources that cannot be pooled. We here consider the regression case and propose a method where a conformal predictor is trained on each data source independently, and where the prediction intervals are then combined into a single interval. We call the approach Non-Disclosed Conformal Prediction (NDCP), and we evaluate it on a regression dataset from the UCI machine learning repository using support vector regression as the underlying machine learning algorithm, with varying number of data sources and sizes. The results show that the proposed method produces conservatively valid prediction intervals, and while we cannot retain the same efficiency as when all data is used, efficiency is improved through the proposed approach as compared to predicting using a single arbitrarily chosen source. Reference: Spjuth O., Brännström R.C., Carlsson L. and Gauraha, N. Combining Prediction Intervals on Multi-Source Non-Disclosed Regression Datasets. Proceedings of Machine Learning Research (PMLR). 105, 53-65. (2019). URL: https://proceedings.mlr.press/v105/spjuth19a.html ]]>
Presentation given by Ola Spjuth at the 8th Symposium on Conformal and Probabilistic Prediction with Applications in Varna, Bulgaria on Sept 20th, 2019. ABSTRACT Conformal Prediction is a framework that produces prediction intervals based on the output from a machine learning algorithm. In this paper we explore the case when training data is made up of multiple parts available in different sources that cannot be pooled. We here consider the regression case and propose a method where a conformal predictor is trained on each data source independently, and where the prediction intervals are then combined into a single interval. We call the approach Non-Disclosed Conformal Prediction (NDCP), and we evaluate it on a regression dataset from the UCI machine learning repository using support vector regression as the underlying machine learning algorithm, with varying number of data sources and sizes. The results show that the proposed method produces conservatively valid prediction intervals, and while we cannot retain the same efficiency as when all data is used, efficiency is improved through the proposed approach as compared to predicting using a single arbitrarily chosen source. Reference: Spjuth O., Brännström R.C., Carlsson L. and Gauraha, N. Combining Prediction Intervals on Multi-Source Non-Disclosed Regression Datasets. Proceedings of Machine Learning Research (PMLR). 105, 53-65. (2019). URL: https://proceedings.mlr.press/v105/spjuth19a.html ]]> Tue, 10 Sep 2019 11:53:36 GMT /slideshow/combining-prediction-intervals-on-multisource-nondisclosed-regression-datasets/170566798 ospjuth@slideshare.net(ospjuth) Combining Prediction Intervals on Multi-Source Non-Disclosed Regression Datasets ospjuth Presentation given by Ola Spjuth at the 8th Symposium on Conformal and Probabilistic Prediction with Applications in Varna, Bulgaria on Sept 20th, 2019. ABSTRACT Conformal Prediction is a framework that produces prediction intervals based on the output from a machine learning algorithm. In this paper we explore the case when training data is made up of multiple parts available in different sources that cannot be pooled. We here consider the regression case and propose a method where a conformal predictor is trained on each data source independently, and where the prediction intervals are then combined into a single interval. We call the approach Non-Disclosed Conformal Prediction (NDCP), and we evaluate it on a regression dataset from the UCI machine learning repository using support vector regression as the underlying machine learning algorithm, with varying number of data sources and sizes. The results show that the proposed method produces conservatively valid prediction intervals, and while we cannot retain the same efficiency as when all data is used, efficiency is improved through the proposed approach as compared to predicting using a single arbitrarily chosen source. Reference: Spjuth O., Brännström R.C., Carlsson L. and Gauraha, N. Combining Prediction Intervals on Multi-Source Non-Disclosed Regression Datasets. Proceedings of Machine Learning Research (PMLR). 105, 53-65. (2019). URL: https://proceedings.mlr.press/v105/spjuth19a.html <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/ndcp-copa2019-190910115337-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation given by Ola Spjuth at the 8th Symposium on Conformal and Probabilistic Prediction with Applications in Varna, Bulgaria on Sept 20th, 2019. ABSTRACT Conformal Prediction is a framework that produces prediction intervals based on the output from a machine learning algorithm. In this paper we explore the case when training data is made up of multiple parts available in different sources that cannot be pooled. We here consider the regression case and propose a method where a conformal predictor is trained on each data source independently, and where the prediction intervals are then combined into a single interval. We call the approach Non-Disclosed Conformal Prediction (NDCP), and we evaluate it on a regression dataset from the UCI machine learning repository using support vector regression as the underlying machine learning algorithm, with varying number of data sources and sizes. The results show that the proposed method produces conservatively valid prediction intervals, and while we cannot retain the same efficiency as when all data is used, efficiency is improved through the proposed approach as compared to predicting using a single arbitrarily chosen source. Reference: Spjuth O., Brännström R.C., Carlsson L. and Gauraha, N. Combining Prediction Intervals on Multi-Source Non-Disclosed Regression Datasets. Proceedings of Machine Learning Research (PMLR). 105, 53-65. (2019). URL: https://proceedings.mlr.press/v105/spjuth19a.html

Combining Prediction Intervals on Multi-Source Non-Disclosed Regression Datasets from Ola Spjuth

]]> 353 0 https://cdn.slidesharecdn.com/ss_thumbnails/ndcp-copa2019-190910115337-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Building an informatics solution to sustain AI-guided cell profiling with high-content microscopy imaging /slideshow/building-an-informatics-solution-to-sustain-aiguided-cell-profiling-with-highcontent-microscopy-imaging/169000944 spjuthslaseurope2019final-190904133032
Presentation at SLAS Europe 2019 in Barcelona on 28 june, 2019. High-content microscopy in automated laboratories present many challenges for storing and processing data, and to build AI models to aid decision making. We have established an informatics system to serve a robotized cell profiling setup with incubators, liquid handling and high-content microscopy for microplates. The informatics system consists of computational infrastructure (CPUs, GPUs, storage), middleware (Kubernetes), imaging database and software (OMERO), and workflow system (Pachyderm) to perform online prioritization of new data, and automate the process from acquired images to continuously updated and deployed AI models. The AI methodologies include Deep Learning models trained on image data, and conventional machine learning models trained on data from Cell Painting experiments. The microservice architecture makes the system scalable and expandable, and a key objective is on improving screening and toxicity assessment using AI-aided intelligent experimental design.]]>
Presentation at SLAS Europe 2019 in Barcelona on 28 june, 2019. High-content microscopy in automated laboratories present many challenges for storing and processing data, and to build AI models to aid decision making. We have established an informatics system to serve a robotized cell profiling setup with incubators, liquid handling and high-content microscopy for microplates. The informatics system consists of computational infrastructure (CPUs, GPUs, storage), middleware (Kubernetes), imaging database and software (OMERO), and workflow system (Pachyderm) to perform online prioritization of new data, and automate the process from acquired images to continuously updated and deployed AI models. The AI methodologies include Deep Learning models trained on image data, and conventional machine learning models trained on data from Cell Painting experiments. The microservice architecture makes the system scalable and expandable, and a key objective is on improving screening and toxicity assessment using AI-aided intelligent experimental design.]]> Wed, 04 Sep 2019 13:30:31 GMT /slideshow/building-an-informatics-solution-to-sustain-aiguided-cell-profiling-with-highcontent-microscopy-imaging/169000944 ospjuth@slideshare.net(ospjuth) Building an informatics solution to sustain AI-guided cell profiling with high-content microscopy imaging ospjuth Presentation at SLAS Europe 2019 in Barcelona on 28 june, 2019. High-content microscopy in automated laboratories present many challenges for storing and processing data, and to build AI models to aid decision making. We have established an informatics system to serve a robotized cell profiling setup with incubators, liquid handling and high-content microscopy for microplates. The informatics system consists of computational infrastructure (CPUs, GPUs, storage), middleware (Kubernetes), imaging database and software (OMERO), and workflow system (Pachyderm) to perform online prioritization of new data, and automate the process from acquired images to continuously updated and deployed AI models. The AI methodologies include Deep Learning models trained on image data, and conventional machine learning models trained on data from Cell Painting experiments. The microservice architecture makes the system scalable and expandable, and a key objective is on improving screening and toxicity assessment using AI-aided intelligent experimental design. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/spjuthslaseurope2019final-190904133032-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation at SLAS Europe 2019 in Barcelona on 28 june, 2019. High-content microscopy in automated laboratories present many challenges for storing and processing data, and to build AI models to aid decision making. We have established an informatics system to serve a robotized cell profiling setup with incubators, liquid handling and high-content microscopy for microplates. The informatics system consists of computational infrastructure (CPUs, GPUs, storage), middleware (Kubernetes), imaging database and software (OMERO), and workflow system (Pachyderm) to perform online prioritization of new data, and automate the process from acquired images to continuously updated and deployed AI models. The AI methodologies include Deep Learning models trained on image data, and conventional machine learning models trained on data from Cell Painting experiments. The microservice architecture makes the system scalable and expandable, and a key objective is on improving screening and toxicity assessment using AI-aided intelligent experimental design.

Building an informatics solution to sustain AI-guided cell profiling with high-content microscopy imaging from Ola Spjuth

]]> 323 1 https://cdn.slidesharecdn.com/ss_thumbnails/spjuthslaseurope2019final-190904133032-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Automating the process of continuously prioritising data, updating and deploying AI models in a robotised lab for drug discovery /slideshow/automating-the-process-of-continuously-prioritising-data-updating-and-deploying-ai-models-in-a-robotised-lab-for-drug-discovery/168993955 spjuthdis2019-190904124933
Presentation at Data Innovation Summit 2019 in Stockholm, Sweden. ABSTRACT Microscopes are capable of producing vast amounts of data, and when used in automated laboratories both the number and size of images present many challenges for storing, categorizing, analyzing, annotating, and transforming the data into actionable information that can used for decision making; either by humans or machines. In this presentation I will describe the informatics system we have established at the Department of Pharmaceutical Biosciences at Uppsala University, which consists of computational hardware (CPUs, GPUs, storage), middleware (Kubernetes), imaging database (OMERO), and workflow system (Pachyderm) to perform online prioritization of new data, as well as the continuous analytics system to automate the process from captured images to continuously updated and deployed AI models. The AI methodologies include Deep Learning models trained on image data, and conventional machine learning models trained on features extracted from images or chemical structures. Due to the microservice architecture the system is scalable and can be expanded using hybrid-architectures with cloud computing resources. The informatics system serves a robotized cell profiling setup with incubators, liquid handling and high-content microscopy. The lab is quite young and is targeting applications primarily in drug screening and toxicity assessment, with the aim to improve research using AI and intelligent design of experiments.]]>
Presentation at Data Innovation Summit 2019 in Stockholm, Sweden. ABSTRACT Microscopes are capable of producing vast amounts of data, and when used in automated laboratories both the number and size of images present many challenges for storing, categorizing, analyzing, annotating, and transforming the data into actionable information that can used for decision making; either by humans or machines. In this presentation I will describe the informatics system we have established at the Department of Pharmaceutical Biosciences at Uppsala University, which consists of computational hardware (CPUs, GPUs, storage), middleware (Kubernetes), imaging database (OMERO), and workflow system (Pachyderm) to perform online prioritization of new data, as well as the continuous analytics system to automate the process from captured images to continuously updated and deployed AI models. The AI methodologies include Deep Learning models trained on image data, and conventional machine learning models trained on features extracted from images or chemical structures. Due to the microservice architecture the system is scalable and can be expanded using hybrid-architectures with cloud computing resources. The informatics system serves a robotized cell profiling setup with incubators, liquid handling and high-content microscopy. The lab is quite young and is targeting applications primarily in drug screening and toxicity assessment, with the aim to improve research using AI and intelligent design of experiments.]]> Wed, 04 Sep 2019 12:49:33 GMT /slideshow/automating-the-process-of-continuously-prioritising-data-updating-and-deploying-ai-models-in-a-robotised-lab-for-drug-discovery/168993955 ospjuth@slideshare.net(ospjuth) Automating the process of continuously prioritising data, updating and deploying AI models in a robotised lab for drug discovery ospjuth Presentation at Data Innovation Summit 2019 in Stockholm, Sweden. ABSTRACT Microscopes are capable of producing vast amounts of data, and when used in automated laboratories both the number and size of images present many challenges for storing, categorizing, analyzing, annotating, and transforming the data into actionable information that can used for decision making; either by humans or machines. In this presentation I will describe the informatics system we have established at the Department of Pharmaceutical Biosciences at Uppsala University, which consists of computational hardware (CPUs, GPUs, storage), middleware (Kubernetes), imaging database (OMERO), and workflow system (Pachyderm) to perform online prioritization of new data, as well as the continuous analytics system to automate the process from captured images to continuously updated and deployed AI models. The AI methodologies include Deep Learning models trained on image data, and conventional machine learning models trained on features extracted from images or chemical structures. Due to the microservice architecture the system is scalable and can be expanded using hybrid-architectures with cloud computing resources. The informatics system serves a robotized cell profiling setup with incubators, liquid handling and high-content microscopy. The lab is quite young and is targeting applications primarily in drug screening and toxicity assessment, with the aim to improve research using AI and intelligent design of experiments. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/spjuthdis2019-190904124933-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation at Data Innovation Summit 2019 in Stockholm, Sweden. ABSTRACT Microscopes are capable of producing vast amounts of data, and when used in automated laboratories both the number and size of images present many challenges for storing, categorizing, analyzing, annotating, and transforming the data into actionable information that can used for decision making; either by humans or machines. In this presentation I will describe the informatics system we have established at the Department of Pharmaceutical Biosciences at Uppsala University, which consists of computational hardware (CPUs, GPUs, storage), middleware (Kubernetes), imaging database (OMERO), and workflow system (Pachyderm) to perform online prioritization of new data, as well as the continuous analytics system to automate the process from captured images to continuously updated and deployed AI models. The AI methodologies include Deep Learning models trained on image data, and conventional machine learning models trained on features extracted from images or chemical structures. Due to the microservice architecture the system is scalable and can be expanded using hybrid-architectures with cloud computing resources. The informatics system serves a robotized cell profiling setup with incubators, liquid handling and high-content microscopy. The lab is quite young and is targeting applications primarily in drug screening and toxicity assessment, with the aim to improve research using AI and intelligent design of experiments.

Automating the process of continuously prioritising data, updating and deploying AI models in a robotised lab for drug discovery from Ola Spjuth

]]> 264 0 https://cdn.slidesharecdn.com/ss_thumbnails/spjuthdis2019-190904124933-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Data-intensive applications on cloud computing resources: Applications in life sciences /slideshow/dataintensive-applications-on-cloud-computing-resources-applications-in-life-sciences/81241686 spjuthdenbiv2-2017-171026125829
Presentation at the de.NBI 2017 symposium “The Future Development of Bioinformatics in Germany and Europe” held at the Center for Interdisciplinary Research (ZiF) of Bielefeld University, October 23-25, 2017. https://www.denbi.de/symposium2017]]>
Presentation at the de.NBI 2017 symposium “The Future Development of Bioinformatics in Germany and Europe” held at the Center for Interdisciplinary Research (ZiF) of Bielefeld University, October 23-25, 2017. https://www.denbi.de/symposium2017]]> Thu, 26 Oct 2017 12:58:29 GMT /slideshow/dataintensive-applications-on-cloud-computing-resources-applications-in-life-sciences/81241686 ospjuth@slideshare.net(ospjuth) Data-intensive applications on cloud computing resources: Applications in life sciences ospjuth Presentation at the de.NBI 2017 symposium “The Future Development of Bioinformatics in Germany and Europe” held at the Center for Interdisciplinary Research (ZiF) of Bielefeld University, October 23-25, 2017. https://www.denbi.de/symposium2017 <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/spjuthdenbiv2-2017-171026125829-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation at the de.NBI 2017 symposium “The Future Development of Bioinformatics in Germany and Europe” held at the Center for Interdisciplinary Research (ZiF) of Bielefeld University, October 23-25, 2017. https://www.denbi.de/symposium2017

Data-intensive applications on cloud computing resources: Applications in life sciences from Ola Spjuth

]]> 445 4 https://cdn.slidesharecdn.com/ss_thumbnails/spjuthdenbiv2-2017-171026125829-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Data-intensive bioinformatics on HPC and Cloud /slideshow/dataintensive-bioinformatics-on-hpc-and-cloud/80109725 spjuthcharme-bigdata-170924211544
My presentation at the COST-CHARME 'Big Data Training School for Life Sciences', on 18-22 September 2017 in Uppsala, Sweden.]]>
My presentation at the COST-CHARME 'Big Data Training School for Life Sciences', on 18-22 September 2017 in Uppsala, Sweden.]]> Sun, 24 Sep 2017 21:15:44 GMT /slideshow/dataintensive-bioinformatics-on-hpc-and-cloud/80109725 ospjuth@slideshare.net(ospjuth) Data-intensive bioinformatics on HPC and Cloud ospjuth My presentation at the COST-CHARME 'Big Data Training School for Life Sciences', on 18-22 September 2017 in Uppsala, Sweden. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/spjuthcharme-bigdata-170924211544-thumbnail.jpg?width=120&height=120&fit=bounds" /> My presentation at the COST-CHARME 'Big Data Training School for Life Sciences', on 18-22 September 2017 in Uppsala, Sweden.

Data-intensive bioinformatics on HPC and Cloud from Ola Spjuth

]]> 1409 8 https://cdn.slidesharecdn.com/ss_thumbnails/spjuthcharme-bigdata-170924211544-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 The case for cloud computing in Life Sciences /ospjuth/the-case-for-cloud-computing-in-life-sciences spjuth-cloudbeer-170330204543
Presentation by Ola Spjuth at CloudBeer 2017 in Stockholm, organized by City Networks.]]>
Presentation by Ola Spjuth at CloudBeer 2017 in Stockholm, organized by City Networks.]]> Thu, 30 Mar 2017 20:45:43 GMT /ospjuth/the-case-for-cloud-computing-in-life-sciences ospjuth@slideshare.net(ospjuth) The case for cloud computing in Life Sciences ospjuth Presentation by Ola Spjuth at CloudBeer 2017 in Stockholm, organized by City Networks. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/spjuth-cloudbeer-170330204543-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation by Ola Spjuth at CloudBeer 2017 in Stockholm, organized by City Networks.

The case for cloud computing in Life Sciences from Ola Spjuth

]]> 765 8 https://cdn.slidesharecdn.com/ss_thumbnails/spjuth-cloudbeer-170330204543-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Storage and Analysis of Sensitive Large-Scale Biomedical Data in Sweden /slideshow/storage-and-analysis-of-sensitive-largescale-biomedical-data-in-sweden/69933366 spjuth-descrsteinbeckebi-161207231308
Presentation at EBI on Aug 11, 2015]]>
Presentation at EBI on Aug 11, 2015]]> Wed, 07 Dec 2016 23:13:08 GMT /slideshow/storage-and-analysis-of-sensitive-largescale-biomedical-data-in-sweden/69933366 ospjuth@slideshare.net(ospjuth) Storage and Analysis of Sensitive Large-Scale Biomedical Data in Sweden ospjuth Presentation at EBI on Aug 11, 2015 <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/spjuth-descrsteinbeckebi-161207231308-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation at EBI on Aug 11, 2015

Storage and Analysis of Sensitive Large-Scale Biomedical Data in Sweden from Ola Spjuth

]]> 569 3 https://cdn.slidesharecdn.com/ss_thumbnails/spjuth-descrsteinbeckebi-161207231308-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Agile large-scale machine-learning pipelines in drug discovery /slideshow/agile-largescale-machinelearning-pipelines-in-drug-discovery/69933334 spjuth-researchsteinbeckebi-161207231138
Presentation held at EBI on Aug 12, 2015]]>
Presentation held at EBI on Aug 12, 2015]]> Wed, 07 Dec 2016 23:11:38 GMT /slideshow/agile-largescale-machinelearning-pipelines-in-drug-discovery/69933334 ospjuth@slideshare.net(ospjuth) Agile large-scale machine-learning pipelines in drug discovery ospjuth Presentation held at EBI on Aug 12, 2015 <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/spjuth-researchsteinbeckebi-161207231138-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation held at EBI on Aug 12, 2015

Agile large-scale machine-learning pipelines in drug discovery from Ola Spjuth

]]> 606 5 https://cdn.slidesharecdn.com/ss_thumbnails/spjuth-researchsteinbeckebi-161207231138-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Enabling Translational Medicine with e-Science /slideshow/enabling-translational-medicine-with-escience/69933288 ecpcescience-academy2015-161207230838
Presentation at Swedish e-Science Academy 2015]]>
Presentation at Swedish e-Science Academy 2015]]> Wed, 07 Dec 2016 23:08:38 GMT /slideshow/enabling-translational-medicine-with-escience/69933288 ospjuth@slideshare.net(ospjuth) Enabling Translational Medicine with e-Science ospjuth Presentation at Swedish e-Science Academy 2015 <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/ecpcescience-academy2015-161207230838-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation at Swedish e-Science Academy 2015

Enabling Translational Medicine with e-Science from Ola Spjuth

]]> 404 7 https://cdn.slidesharecdn.com/ss_thumbnails/ecpcescience-academy2015-161207230838-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Continuous modeling - automating model building on high-performance e-Infrastructures /slideshow/continuous-modeling-automating-model-building-on-highperformance-einfrastructures/69933166 spjuthicpb-161207230147
Presentation at International Conference on Pharmaceutical Bioinformatics 2016]]>
Presentation at International Conference on Pharmaceutical Bioinformatics 2016]]> Wed, 07 Dec 2016 23:01:47 GMT /slideshow/continuous-modeling-automating-model-building-on-highperformance-einfrastructures/69933166 ospjuth@slideshare.net(ospjuth) Continuous modeling - automating model building on high-performance e-Infrastructures ospjuth Presentation at International Conference on Pharmaceutical Bioinformatics 2016 <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/spjuthicpb-161207230147-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation at International Conference on Pharmaceutical Bioinformatics 2016

Continuous modeling - automating model building on high-performance e-Infrastructures from Ola Spjuth

]]> 329 4 https://cdn.slidesharecdn.com/ss_thumbnails/spjuthicpb-161207230147-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Analyzing Big Data in Medicine with Virtual Research Environments and Microservices /slideshow/analyzing-big-data-in-medicine-with-virtual-research-environments-and-microservices/69929854 spjuthbigdata-medicine-161207205147
Presented at Big Data in Medicine, Uppsala, Sweden on 2016-11-17]]>
Presented at Big Data in Medicine, Uppsala, Sweden on 2016-11-17]]> Wed, 07 Dec 2016 20:51:47 GMT /slideshow/analyzing-big-data-in-medicine-with-virtual-research-environments-and-microservices/69929854 ospjuth@slideshare.net(ospjuth) Analyzing Big Data in Medicine with Virtual Research Environments and Microservices ospjuth Presented at Big Data in Medicine, Uppsala, Sweden on 2016-11-17 <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/spjuthbigdata-medicine-161207205147-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presented at Big Data in Medicine, Uppsala, Sweden on 2016-11-17

Analyzing Big Data in Medicine with Virtual Research Environments and Microservices from Ola Spjuth

]]> 1600 7 https://cdn.slidesharecdn.com/ss_thumbnails/spjuthbigdata-medicine-161207205147-thumbnail.jpg?width=120&height=120&fit=bounds presentation 000000 http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Interoperability and scalability with microservices in science /slideshow/interoperability-and-scalability-with-microservices-in-science/67940707 spjuthopentoxeuro2016-161031215813
Microservices have emerged as a modern interpretation of service-oriented architectures where processes are small and communicate over a network using lightweight protocols to fulfill a goal. In this talk I will present our work on microservices, and how they can be used to empower interoperable and scalable analysis services and pipelines in virtual infrastructures on cloud computing resources. I will also give examples and experiences from the PhenoMeNal H2020 project where a developer community in metabolomics is moving to such architecture.]]>
Microservices have emerged as a modern interpretation of service-oriented architectures where processes are small and communicate over a network using lightweight protocols to fulfill a goal. In this talk I will present our work on microservices, and how they can be used to empower interoperable and scalable analysis services and pipelines in virtual infrastructures on cloud computing resources. I will also give examples and experiences from the PhenoMeNal H2020 project where a developer community in metabolomics is moving to such architecture.]]> Mon, 31 Oct 2016 21:58:13 GMT /slideshow/interoperability-and-scalability-with-microservices-in-science/67940707 ospjuth@slideshare.net(ospjuth) Interoperability and scalability with microservices in science ospjuth Microservices have emerged as a modern interpretation of service-oriented architectures where processes are small and communicate over a network using lightweight protocols to fulfill a goal. In this talk I will present our work on microservices, and how they can be used to empower interoperable and scalable analysis services and pipelines in virtual infrastructures on cloud computing resources. I will also give examples and experiences from the PhenoMeNal H2020 project where a developer community in metabolomics is moving to such architecture. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/spjuthopentoxeuro2016-161031215813-thumbnail.jpg?width=120&height=120&fit=bounds" /> Microservices have emerged as a modern interpretation of service-oriented architectures where processes are small and communicate over a network using lightweight protocols to fulfill a goal. In this talk I will present our work on microservices, and how they can be used to empower interoperable and scalable analysis services and pipelines in virtual infrastructures on cloud computing resources. I will also give examples and experiences from the PhenoMeNal H2020 project where a developer community in metabolomics is moving to such architecture.

Interoperability and scalability with microservices in science from Ola Spjuth

]]> 1571 6 https://cdn.slidesharecdn.com/ss_thumbnails/spjuthopentoxeuro2016-161031215813-thumbnail.jpg?width=120&height=120&fit=bounds presentation 000000 http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Chemical decision support in toxicology and pharmacology (OpenToxEU 2013) /slideshow/bioclipse-at-opentox-euro-2013/27407547 bioclipseopentox2013-noscreencasts-131021063040-phpapp02
Presentation given at OpenTox Euro 2013.]]>
Presentation given at OpenTox Euro 2013.]]> Mon, 21 Oct 2013 06:30:40 GMT /slideshow/bioclipse-at-opentox-euro-2013/27407547 ospjuth@slideshare.net(ospjuth) Chemical decision support in toxicology and pharmacology (OpenToxEU 2013) ospjuth Presentation given at OpenTox Euro 2013. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/bioclipseopentox2013-noscreencasts-131021063040-phpapp02-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation given at OpenTox Euro 2013.

Chemical decision support in toxicology and pharmacology (OpenToxEU 2013) from Ola Spjuth

]]> 1222 4 https://cdn.slidesharecdn.com/ss_thumbnails/bioclipseopentox2013-noscreencasts-131021063040-phpapp02-thumbnail.jpg?width=120&height=120&fit=bounds presentation 000000 http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Building a flexible infrastructure with Bioclipse, open source, and federated cloud services /slideshow/building-a-flexible-infrastructure-with-bioclipse-open-source-and-federated-cloud-services/3363913 bioit09-sent-100308054946-phpapp02
Presentation held at Bio-IT World Expo Europe 2009. Presenters: * Ola Spjuth, Dept. Pharmaceutical Biosciences, Uppsala University, Sweden * Lars Carlsson, Global Safety Assessment, AstraZeneca R&D, Sweden]]>
Presentation held at Bio-IT World Expo Europe 2009. Presenters: * Ola Spjuth, Dept. Pharmaceutical Biosciences, Uppsala University, Sweden * Lars Carlsson, Global Safety Assessment, AstraZeneca R&D, Sweden]]> Mon, 08 Mar 2010 05:49:43 GMT /slideshow/building-a-flexible-infrastructure-with-bioclipse-open-source-and-federated-cloud-services/3363913 ospjuth@slideshare.net(ospjuth) Building a flexible infrastructure with Bioclipse, open source, and federated cloud services ospjuth Presentation held at Bio-IT World Expo Europe 2009. Presenters: * Ola Spjuth, Dept. Pharmaceutical Biosciences, Uppsala University, Sweden * Lars Carlsson, Global Safety Assessment, AstraZeneca R&D, Sweden <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/bioit09-sent-100308054946-phpapp02-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation held at Bio-IT World Expo Europe 2009. Presenters: * Ola Spjuth, Dept. Pharmaceutical Biosciences, Uppsala University, Sweden * Lars Carlsson, Global Safety Assessment, AstraZeneca R&D, Sweden

Building a flexible infrastructure with Bioclipse, open source, and federated cloud services from Ola Spjuth

]]> 683 5 https://cdn.slidesharecdn.com/ss_thumbnails/bioit09-sent-100308054946-phpapp02-thumbnail.jpg?width=120&height=120&fit=bounds presentation 000000 http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Accessing and scripting CDK �from Bioclipse /slideshow/accessing-and-scripting-cdk-from-bioclipse/3363912 cdkws-spjuth3-100308054954-phpapp02
Presentation held at the CDK workshop 2009.]]>
Presentation held at the CDK workshop 2009.]]> Mon, 08 Mar 2010 05:49:37 GMT /slideshow/accessing-and-scripting-cdk-from-bioclipse/3363912 ospjuth@slideshare.net(ospjuth) Accessing and scripting CDK �from Bioclipse ospjuth Presentation held at the CDK workshop 2009. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/cdkws-spjuth3-100308054954-phpapp02-thumbnail.jpg?width=120&height=120&fit=bounds" /> Presentation held at the CDK workshop 2009.

Accessing and scripting CDK from Bioclipse from Ola Spjuth

]]> 1130 5 https://cdn.slidesharecdn.com/ss_thumbnails/cdkws-spjuth3-100308054954-phpapp02-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 https://cdn.slidesharecdn.com/profile-photo-ospjuth-48x48.jpg?cb=1679224244 Associate professor in Bioinformatics from Uppsala University. Research interest: High-performance bioinformatics, computational pharmacology, and research on e-Science infrastructures for life science research. Involved in the startup company Genetta Soft (www.genettasoft.com). pharmb.io https://cdn.slidesharecdn.com/ss_thumbnails/spjuthelrig-roboautov3-221208082233-ef31cf5e-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/automating-cellbased-screening-with-open-source-robotics-and-ai/254822533 Automating cell-based ... https://cdn.slidesharecdn.com/ss_thumbnails/chembioseminarki-200207191519-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/towards-automated-phenotypic-cell-profiling-with-highcontent-imaging/227300820 Towards automated phen... https://cdn.slidesharecdn.com/ss_thumbnails/spjuthescienceacademy2019-v2-191017142124-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/towards-automated-aiguided-drug-discovery-labs/183154246 Towards Automated AI-g...